Microsoft Word - It's time to admit TVAAS is broken.docx

By Sally Absher

Kudos to the Knox County Schools Board of Education, for accepting a waiver from the state to not include the TNReady Quick Scores in student grades this year. Unfortunately, there is no waiver from including the test scores, as part of Tennessee Value Added Assessment System (TVAAS), in teacher evaluations and school rankings. And that is a problem.

The state just released the 2014-2015 “Report Card,” available at After last week’s column on teachers who resigned from Bearden High and South Doyle Middle, both Level 1 schools based on TVAAS scores in 2014-2015, we were curious to see how other high schools performed on TVAAS.

TVAAS is a statistical analysis used to measure the impact of districts, schools, and teachers on the academic progress of groups of students from year to year. It estimates, or “predicts” the growth a group of students will have from one year to the next. TVAAS scores are given between 1 and 5. Levels 4 and 5 indicate that a school is exceeding the expected growth, Level 3 indicates that they are making about the expected growth, and Levels 1 and 2 indicate they are making less than the expected growth. Level 1 schools are “failing schools” and can be taken over by the state if they don’t improve.

This August, Knox County Schools had a big celebration, complete with confetti and balloons, because KCS was ranked an “exceptional” school system – TVAAS score of 5’s across the board, all A’s on achievement, etc. But the scores from individual schools tell a different story. TVASS Composite scores are available for every public school in Tennessee on the Report Card page at

We examined the 2014-2015 overall TVAAS score for 16 KCS High Schools, including the Career and Technical Magnet (CTE) and Kelly Volunteer Academy. Four received overall TVAAS scores of 5: Carter, Farragut, Halls, and Powell. Hardin Valley Academy had an overall TVAAS score of 4. Austin East and Karns had a TVAAS score of 3, and CTE Magnet and Fulton had scores of 2. A total of six KCS high schools had composite TVAAS scores of 1: Bearden, Central, Kelley Volunteer Academy, Gibbs, L&N Stem Academy, and West.

We also compared the 2014-2015 data to TVAAS scores for these high schools from the previous two years – 2012-2013, and 2013-2014. The accompanying table shows the comparison. The first number in each grouping (e.g., 1/2/3/4) is the overall TVAAS composite number.

While some schools have vacillated up and down since 2013, the number of high schools with TVAAS composite scores of 5 has fallen from nine, to four. The number of high schools with TVAAS composite scores of 1 has risen from two, to six. Why is no one calling for the elimination of this junk science of “predicted growth scores” used to evaluate teachers and schools?

After last week’s column, the Focus received an email from a KCS teacher, who wrote:

“I know teachers at Bearden, my two children graduated from that high school.  Right now, the school system is going to spend a tremendous amount of money (coaching, loss of planning/grading time) to solve a problem that doesn’t exist.  No one is going to ask the question:  How can a school go from “5” to “1” in a single year?  Same books.  Same teachers.  Same standards.  Same students.

The problem is Value Added.  Value added predicts what a student should make on the End of Course test.  How it does this, I don’t know.

I taught Algebra 2 for a number of years, at two different high schools.  We study data sets in that course.  One of the problems in the county-issued textbook has students calculate the average for a data set.  Then it asks them to look at the data to see if there is anything odd about them.  The students recognize two outliers.  They recalculate an average without the outliers, and the average goes way up.

Value Added does not do this.  It counts outliers. This is bad math.

I have personal experience with the negatives of Value Added.   There are teachers all over the county that do. The problem with this entire evaluation system is that there is no step, no place, no point in the process where someone can say “after looking at all of the evidence, we conclude that the problem is not the teacher.””

Likewise, how can a teacher go from a “5” to a “1” in a single year? What happens to non-tenured teachers at a Level 1 school who now have a “1” for value added this year? Will they earn tenure if they are otherwise eligible, or will they be non-renewed?

What can be done? Parents and teachers complain to the District, the District blames the State, and the State blames the U.S. Department of Education. Tennessee is known for having one of the most draconian educational policies mandating high-stakes use of assessment tests across the nation.

According to former state Representative Gloria Johnson, the mandated “accountability system” is actually state and local. The state requires it based on their “interpretation” of federal law, but the local District can choose a system. While some changes in the law need to happen at the state level, a smart superintendent could develop a process designed by teachers that made sense and met the mandate…but then a good evaluation of students or teachers is not really the goal.