What accounts for teacher quality? A school’s grade provides only part of the answer.

Four years ago, I helped create and lead a school improvement and professional development partnership between the University of Florida’s Lastinger Center for Learning and the Pinellas County school district. As part of this project I conducted 3,500 classroom observations over two years in which I assessed the quality of student and teacher engagement when I entered each classroom. Second-graders engaged in a turn and talk with an evaluation or synthesis prompt would score high, second-graders sitting at their desks doodling on a worksheet while their teacher shopped online at her desk would score low, and second-graders sitting passively at their desks while their teacher talked would fall in the middle. While this assessment was not a comprehensive measure of instructional quality, it provided a good snapshot and, with 3,500 data points, patterns were easy to discern.

The first year, each school’s data fell into a bell curve, with about 10 percent to 15 percent of the instruction falling at the top and bottom of the curve and 70 percent to 80 percent falling in the middle. Apparently teacher-centric instruction is still the norm in our schools, just as it’s been for the last 200 years. Toward the end of year one, we presented our findings to the schools and provided some professional development. Consequently the year two data skewed more positively with most of the low-end assessments moving to the middle. We saw only a slight increase in the upper levels.

I was unaware of each school’s state grade when I made my observations, but given the negligible variation between schools that I found in either year, I decided to see if there was a relationship between school grades and quality of teaching. There was none. That is, the quality of student and teacher engagement in schools graded A was identical to that in schools graded B, C and D. There were no F schools in our sample.

But while there was no variation between schools, the variation within schools was large. Every school had a small number of teachers who consistently scored high on our scale and an equally small number who consistently scored low.

I was skeptical about Florida’s school grading system when it was implemented, but the benefits have been undeniable. Schools and school districts have focused more resources on low-income and minority students and, as a result, these student populations have seen significant improvements in their standardized test scores. But my research suggests Florida’s school grades do not reflect differences in teacher quality between schools. Instead differences in test scores seems to be caused by differences in student demographics and school leadership.

Even though instructional quality may be a constant across schools, talented school leaders can manage their instructional resources such that the neediest children receive the best instruction, and many do. Unfortunately Florida’s constitutional class-size caps have severely limited school leaders’ flexibility.

The bottom line is that, in the world of customized learning, teacher choice is far more important than school choice.