William Sanders, the Tennessee statistician and researcher who came up with the nation’s first system for evaluating teachers based on student growth, kicking off a contentious, decades-long debate about how best to measure learning, has died.

Sanders died late last week of natural causes in a hospital in Franklin, Tenn., his family said. He was 74.

A former professor at the University of Tennessee and senior research fellow with the University of North Carolina at Chapel Hill, Sanders is best known as the developer of the Tennessee Value-Added Assessment System, or TVAAS. That model has become the foundation for judging the effectiveness of teachers in Tennessee public schools, and has been emulated in North Carolina, Ohio, Pennsylvania and cities across the nation — playing a key role in one of education reform’s central debates.

Tennessee Education Commissioner Candice McQueen called Sanders’ death “a loss to the education community.”

“During his career, Dr. Sanders made significant contributions to the conversation on how to distinguish our most effective educators in terms of improving academic achievement,” McQueen said in a statement on Monday.

Sanders’ value-added model, also known as the Educational Value-Added Assessment System, became a lightning rod for criticism by many teachers and teachers unions skeptical about whether it yields fair and unbiased estimates.

The model has prompted numerous federal lawsuits charging that the evaluation system, which is now tied to teacher pay and tenure in Tennessee, doesn’t take into account student-level variables such as growing up in poverty. In 2014, the American Statistical Association called its validity into question, and other critics have said TVAAS should not be the sole tool used to judge teachers.

The method measures the effects of a teacher, school or district on student performance by tracking the progress of students against the progress they would be expected to make based on their previous performance. The formula is complex, the method requires a huge database, and the name is a mouthful to say. But the model is meant to show the “value” that was added by each teacher, school or district when measured by the change in student test scores each year.

Sanders’ research soon zeroed in on teachers as the most important part of the equation.

“Determining the effectiveness of individual teachers hold the most promise because, again and again, findings from TVAAS research show teacher effectiveness to be the most important factor in the academic growth of students,” he co-wrote in a 1998 paper. “A component linking teacher effectiveness to student outcomes is a necessary part of any educational evaluation system.”

Sanders went on to become a national leader in policy discussions on value-added assessments.

In his obituary, his family said that Sanders’ findings challenged decades of assumptions that the impact of student family life, income or ethnicity superseded the quality of classroom instruction. That conclusion has been complicated by other research showing that teachers matter more than other aspects of a school, but not as much as outside factors like poverty.

Sanders “stood for a hopeful view that teacher effectiveness dwarfs all other factors as a predictor of student academic growth,” his family said. He believed that “educational influence matters and teachers matter most.”

Growing up on a Tennessee dairy farm, Sanders devoted most of his research to agricultural or wildlife questions at the University of Tennessee until 1981, when he came across a newspaper article suggesting that there was no way to hold teachers accountable based on test scores. He disagreed and wrote the office of then-Gov. Lamar Alexander to say that the effectiveness of teaching could be measured against the rate of student progress.

“Basically, all I was trying to do is [say] here’s the statistical methodology that solves the problem that some of the critics are talking about,” he told Nashville Public Radio in 2014.

The Tennessee Department of Education commissioned his first wave of research beginning in 1982, and Sanders began by looking at student and teacher data in Knox County. He found that he could measure the impact that a teacher had on a student’s trajectory if he tracked that student’s data over time.

The resulting TVAAS methodology linked student academic outcomes to educational evaluation for the first time. Tennessee teachers began using the data in 1997, and their evaluations became tied to the tool under a 2010 state law.

While teachers and teachers unions pushed back, state lawmakers followed the urging of then-Gov. Phil Bredesen, who said changing the way teachers are evaluated would help the state win a $500 million Race to the Top grant, which Tennessee went on to receive.

TVAAS made Sanders a target for some teachers, who felt like he didn’t understand their work and created a system that was used against them unfairly. But colleagues remembered him as a teacher himself who cared about teachers and students.

“Pennsylvania has Bill to thank for changing the conversations about students — from why they can’t achieve to discussions about growing student[s] at all levels,” said Kristin Lewald, who spearheaded the TVAAS counterpart in that state.