Debating School Performance Measures

How Should We Measure School Performance?

As BESE gets ready to vote later this month on a new school accountability model¬†for Louisiana, one key issue has been how much weight should be given to student growth in determining a school’s letter grade. This question has generated a lot of discussion and contention.

The superintendent and the Accountability Commission have recommended that student growth count for 25% of the school performance score for K-8 schools and 12.5% for high schools. Nine education and business groups criticized this recommendation, saying it would over-emphasize student progress and could mislead parents and the public about a school’s performance. In a guest column in the Advocate, Michael Petrilli of the Fordham Institute disagreed with their assessment, saying academic growth is actually a more accurate reflection of teacher and school performance.

Why the debate?

At a high level, there are two ways to use test results to judge a school. One is by status: How do students perform at a particular moment in time. The other is growth: How well has the school improved student performance over the course of the year. Both measurements are valid, but they measure different things.

If we are using the letter grade to judge the performance of the adults in the building or to motivate adults to change their behavior, growth is the better measure. How strong was the teaching? How much did students learn year over year? However, teachers are also motivated by the absolute status: Are students prepared to be successful in the next grade level?

If we are using letter grades to communicate to parents the quality of a school, then both status and growth matter. Parents want to know how well students in a school are performing in absolute terms, and they also want to know how much a school is contributing to their student’s growth over time.

If we are using letter grades to communicate to the general public the performance of the school, then status is the most important: How well did students do on the tests? Are they college and career ready?

How do other states balance status versus growth?

Former Florida Governor Jeb Bush was a key influencer in pushing Louisiana to move from performance labels to letter grades. Under his leadership, Florida had strong gains in student performance and saw its ranking on the national NAEP tests rise impressively. Florida gave schools one letter grade, but the score was calculated using absolute student performance (status) for 50% of the weight, and growth/value-added measures for the other 50%.

In fact, 32 states currently include student growth in their K-8 accountability systems, ranging from 20% to 75%, with an average weight of 40%.

Why are so many states including growth?

Let’s look at Apple Elementary School. At Apple Elementary,¬†more than 90% of the students are from low income families, and most come from single parent households. Some parents are high school dropouts, and few parents have any post-secondary education. Many of Apple’s kindergarteners come to school not knowing their letters and with a very limited vocabulary. They begin school far behind their more affluent peers.

At Apple Elementary, the teachers are exceptional and their students are learning the equivalent of 1.5 years of instruction during the school year (think effect size or value added). However, if the school is judged using the status model – how students performed on the day of testing – Apple will likely earn a D or F, no matter how exceptional the teaching, for their students entered school so far behind.

Giving Apple a D or F letter grade does not reflect the high quality of teaching at the school. On the other hand, if Apple Elementary were judged only on growth, it would likely have an A, which would not accurately reflect that most students in the school are performing below grade level.

So, what grade should Apple Elementary get?

Based on the recommendations that will be presented to BESE later this month, Apple Elementary would likely improve one letter grade (F to D or D to C) if growth is weighted 25%. One might argue the school should get more of a bump, but at a minimum Apple should not be a failing school. Nor should it be targeted for intervention, for it is doing an excellent job with its harder to serve student population.

Ideally, Apple Elementary would get two grades, one accurately reflecting status (most students are performing below grade level) and one reflecting growth (the teachers are doing an outstanding job given their student population). However, in 2010, the legislature passed a law requiring BESE to give schools only one score/letter grade. So, BESE must now decide how much weight to give student growth.

How should we measure school performance?

Educate Now! strongly disagrees with those who suggest that the proposed 25% weight for student growth would “inflate school scores” and mislead parents about a school’s performance. The fact is, two schools can have the same performance on the tests and differ greatly in quality. One school has students who started farther behind and did a great job in helping them catch up; the other stared with higher performing students and had mediocre teaching with little student growth. Giving them the same grade is misleading to parents, as Ed Navigator points out in this excellent post.

At a minimum, the new accountability model should include 25% for growth (if not more), but there should be transparency in how this information is reported to families and the public. By law, the Department of Education can only issue one letter grade for each school, but the Department also issues annual School Report Cards for every school. These report cards provide much more detailed information about a school and its performance.

Updating the School Report Cards to include separate grades for the student growth component and the status component would be an excellent way to maintain transparency and give parents the information they need to make informed choices.