Unleashing the Disruptive Power of Data: An L.A. Story
If you want to understand why it is critical to to use data for improving the quality of teaching and leadership in American public education, just consider Shirley Avenue Elementary in the Southern California suburb of Reseda.
At first glance, Shirley Avenue seems to be doing well. After all, the school, which has an 83-percent Latino population (and with 37 percent of its students considered English Language Learners) has made Adequate Yearly Progress for the past four years. All of its students, especially its poor students (who make up almost all of the student population) are reaching proficiency as set by California state law.
But a closer look at the performance of the school’s third and fifth grade teachers — courtesy of the Los Angeles Times‘ recently-updated database on teachers at Shirley Avenue and other schools operated by the Los Angeles Unified School District reveals stark differences in the quality of teaching different students can receive each and every day. If anything, if Value-Added data was used as part of AYP, Shirley Elementary would not likely make AYP. In fact, it would likely ranked one of the district’s least-effective schools largely because many of the teachers who have worked at the school over the past seven years haven’t been up to the task. And an even closer look shows sharp variations in instructional quality.
A fifth-grader who was in recent arrival Kathryn Nickerson’s class during the 2009-2010 school years will have lucked out — and chances are, so will the fifth-graders to follow. She is among Shirley Avenue’s — and L.A. Unified’s — most-effective teachers in mathematics, and one of the more-effective English instructors compared to her peers. And this is not just true based on the Times‘ own Value-Added model; her performance remains the same on the three other models used in the database to measure teacher quality. On the other hand, a fifth-grader in the classroom of Nickerson’s longer-serving colleague, Cindy Hillas, wouldn’t fare so well. Hillas is ranked as one of the district’s least-effective teachers in math, both at Shirley Avenue and the district, according to the Times database, and one of the “less-effective” English instructors. And this remains true no matter no matter the Value-Added model used.
Between Nickerson and Hillas is Mary Sahagian, who is on the academic Mendoza line in both subjects; while she fares slightly better on some Value-Added measures than on the one used by the Times, she is still barely treading water. But at least she was better for her students than Edward Goodman, who taught at the Shirley Avenue from the 2005-2006 to 2008-2009 school years. Goodman was not only one of the least-effective fifth-grade teachers in all subjects at that school, he wasn’t exactly doing so well at Liggett Elementary in Pamorama City, Calif., where he served for the previous two years.
For Shirley Avenue’s fourth-graders, things were better. A child taught by Mark G. Gendernalik, a longtime teacher at the school, is getting one of the school’s top-flight instructors in both English and math for the past seven year; he has outranked his peers, according all four Value-Added model found in the Times‘ database. That child’s fourth-grade peer in the class of Gendernalik’s colleague, Rose Cavanagh (also a longtime teacher at the school) will certainly progress nicely in English; she is one of Shirley Avenue’s (and L.A. Unified’s) most-effective reading instructors. But in math, Cavanagh only gets average progress, based on the Times‘ model (although, she does fare better on two of the three other Value-Added models used by the database). For a student who needs a stronger math instructor,a child may be better off with Gendernalik than with Cavanagh. A kid with intense reading needs could benefit from tutelage by either instructor.
And then there is Paul Wainess, who taught both second and third grade at Shirley Avenue in 2010-2011, who is a triple-threat of sorts. Not only has he proven that he can handle improve the performance of seven classes of third graders, he is even capable of teaching kids in second grade. Wainess could be a model for his fellow instructors both at Shirley Avenue and throughout the entire L.A. Unified system. Depending on his other skills (and his own capacity for honing his potential in leadership and social entrepreneurism), he could either be a master instructor or even be placed in the administrative ranks or start a program in which he helps improve student performance.
Certainly the Times database on Shirley Avenue isn’t complete. Data for the 2010-2011 school year isn’t available, nor is there test data on science performance, which are given in fifth grade. But the data available points out something that other research is now starting to show: That the quality of an education within a school can vary from classroom to classroom, and even within the classroom, vary from one subject to another. In some cases, a teacher who is really strong in reading instruction may be a terrible math tutor, while her colleague down the hall is top-notch in that subject. In other cases, a child who falls behind under one teacher while in fourth grade could easily get back on track if the child gets a better teacher in the next grade, and vice versa.
Imagine what a talented principal with strong leadership skills and the ability to hire and fire teachers could do with such data. And it goes beyond simply removing laggard instructors. One possibility: Creating collaborative teaching structures in which teachers strong in reading can handle such instruction for an entire fifth-grade group, while peers talented in math teaching can handle those activities. Another could be that L.A. Unified rewards its top teachers both with bonuses and with recognition of their progress over time. Holding up top teachers as sterling examples of what can be done in classrooms — and even looking at how they do their work — can ultimately help improve the district’s recruitment and training efforts. Meanwhile the data is incredibly useful for parents and families, who should have this information in their arsenal. They can use this data (along with other information) to demand better for their children.
Just imagine the possibilities. But that’s only if the data is available and used.
L.A. Unified has finally begun using Value-Added in both ranking schools and in teacher evaluations; it took the Times‘ series to force the district to do what should have been done long ago. (The district is still playing the political game, with new Superintendent John Deasy joining other L.A. players in demanding that the Times stop publishing information on teacher performance.) States are just starting to now require the use of student data in teacher evaluations. But this is still in infancy, and more importantly, Value-Added data is watered down as sort of a “multiple measure” that renders it useless; we may still have shoddy evaluations instead of rigorous tools that can actually help teachers, principals, families and students. And though the National Education Association and the American Federation of Teachers have conceded some ground on using objective data in evaluations, they are still largely opposed to it; other defenders of traditional public education continue to argue that Value-Added data (and the underlying test data used) are too imperfect for use and doesn’t control for socioeconomic effects in spite of growing evidence to the contrary (and the reality that nothing in the hands of humans will never be perfect).
What status quo defenders should realize is that more-objective data (including other wide-ranging information) can actually improve the quality of education for all children. Besides tossing out laggard instructors, it can also help principals better-use the strengths of their teachers to help students succeed. It will also help break the antagonistic, 20th-century unionism-based model of employee-manager relations that has long dominated schools. By providing principals and teachers with objective performance data (and breaking with the use of less-than-objective evaluations), principals can actually manage teachers instead of simply hoping to move out lemons. It can also promote more-collaborative school environments in which teachers work together.
And for parents, Value-Added data gives them the ability to actually become consumers and lead decision-makers in education. They can use the data to spur overhauls of traditional schools and districts, or decide to use school choice options and send their kids to better-quality options. They can even shop within schools, sending their kids to the better teachers in the building — and signal to principals that change must come. Such parent power is critical to helping everyone — even good-to-great teachers who deserve better than to work with low-quality performers.
When a third of all fourth-graders nationwide are functionally illiterate and 150 high schoolers drop out into prison and poverty, there is no way we can ignore the disruptive potential power of data. It must be harnessed and used to improve the quality of education for every child no matter where they live. And it’s high time to get going.