The Delta Issue #96

A Formula for School Improvement

Hi y’all, Jessica here.

A few weeks back, we launched a school improvement series with the goal of shining a brighter light on what it takes, at its most fundamental level, to improve schools.

This week, we’re honing in on accountability formulas—the often-mysterious mathematical recipes states use to turn raw data about a school into the ratings educators, families, policymakers, and communities see.

Like assessment, the formula is another lever states can pull to influence what kids experience in classrooms. Every weighting decision, every performance threshold, and every reporting deadline tells educators what the state values and where they should focus their attention.

One mistake states make is treating accountability formulas as math problems when, in reality, they are policy decisions. When formulas become overly complicated, focus on the wrong data, or deliver results half a year after the fact, they stop doing the thing they’re supposed to do: help people make better decisions.

A good formula doesn’t just sort schools into categories or identify the lowest performers. It’s how you communicate to every school, regardless of where they fall, what excellence looks like and what to prioritize. 

So if you’re a state leader looking to improve schools, here are three ways to start with the formula.

1️⃣ Build from A-to-F, Not F-to-A.

Federal law requires states to identify the 5 percent of lowest-performing schools, so policymakers naturally spend enormous amounts of time discussing how to identify failure.

This is backwards. When a state designs its formula focusing on identifying the worst schools, the thresholds, weights, and calculations all get calibrated to distinguish the most struggling schools, rather than defining what excellence looks like for everyone else. You end up with a system that is highly focused on the bottom 5 percent but has very little to say to the other 95 percent of schools.

As a result, a school performing in the middle of the pack—not failing, but not exceptional either—looks at its rating, sees it’s safely above the danger zone, and moves on. It learns very little about what to do differently.

There’s a better way: start by defining excellence. Establish what should be true for students in a school earning your state’s highest rating, whether that’s an A, five stars, or some other designation.

When states begin with excellence rather than failure, several things change.

1. First, schools gain clarity about the destination. Educators should know what they’re working toward, not just what they’re trying to avoid. “Don’t be in the bottom 5 percent” is more of a warning label than a roadmap. “Here’s what an A school does for kids” is something a principal and staff can plan around.
2. Second, accountability becomes relevant to every school, not just those at the bottom. A school that’s currently a C now has a clear, concrete picture of what it would take to become a B or an A, not just reassurance that it’s not at risk of intervention. The formula is no longer speaking only to the lowest-performing schools; it’s providing direction for the vast majority of schools that fall somewhere in the middle.
3. Third, the system reinforces a statewide north star. A formula built around excellence communicates what the state believes every student deserves, regardless of zip code or starting point, and creates a consistent set of incentives pulling every school toward that same vision. Identification of the schools most in need of support still happens, but it happens as a natural byproduct of measuring progress toward excellence, not as the organizing principle of the entire system.

2️⃣ Make the Formula Understandable.

Many accountability systems have become so convoluted that the people expected to respond to them cannot understand the basics of how they work.

The complexity is usually well-intentioned. States want accountability systems that are fair, precise, and statistically reliable, and measurement experts are rightly focused on getting the methodology right.

But accountability systems aren’t built for statisticians. They’re built for educators. 

This doesn’t mean states should wholly avoid sophisticated measures such as student growth percentiles or value-added models. These growth models are more complex, but they also provide valuable information unavailable in a more simplistic value-table approach to growth, for example. 

The goal isn’t to eliminate complexity. It’s to be selective about it. States should protect the integrity of the formula by including measures they truly value and avoiding the temptation to add so many indicators with so many permutations nobody can remember what the formula is actually trying to accomplish.

At the end of the day, regardless of the measures chosen, a good accountability system should always pass the principal test: principals should be able to explain how their rating was determined and what they need to do to improve it.

3️⃣ Keep the Formula Connected to Kids.

At the end of the day, accountability ratings are supposed to tell us something about how students are doing. Your formula should make it possible to connect a school’s rating back to the students and outcomes that produced it.

In Louisiana, we intentionally designed a system where accountability scores could be traced back to individual students. You could calculate a score for a school, a grade level, or even a particular classroom. A teacher could look at their roster and see, in concrete terms, how their students were contributing to the school’s performance. A principal could sit with a grade-level team and work backward from the school’s rating to understand exactly which students and which outcomes were driving it.

The goal wasn’t to make the formula simplistic. It was to make it transparent, so the people responsible for improving outcomes could understand what was driving their results and what actions would improve them.

Let’s Get Muddy

• Check out the rest of our school improvement series:
• If you’re heading to Austin next week for Council of Chief State School Officers National Conference on Student Assessment, keep an eye out for Watershed’s Jill Zimmerman Pinsky . She’ll be leading several sessions, including one on how measurement and accountability can support school improvement.
• Want to go deeper on accountability and measurement? Check out Watershed’s Measurement Playbook, authored by Jill Pinsky.