Stay Up-to-Date:
RELevant

Viewpoints and Findings from the REL Mid-Atlantic

The future of school accountability: Not set in stone
By Brian Gill, REL Mid-Atlantic Director

The future of school accountability: Not set in stone

When Congress passed the Every Student Succeeds Act (ESSA) in 2015, one of its aims was to restore some state authority over school accountability systems, in response to criticism that federal requirements had become too prescriptive . Notably, ESSA permits states to choose among new measures of school performance or student outcomes to complement existing measures of reading and math proficiency. Although every state now has an approved ESSA accountability plan, for many states, these plans are refined over time. The school accountability landscape is not set in stone.

In addition to serving as director of REL Mid-Atlantic, I also lead the REL’s research alliance on accountability in the ESSA era. The alliance includes stakeholders from all five mid-Atlantic states who design, implement, and evaluate school accountability systems. Members engage in collaborative technical support and research projects that help refine these systems and their associated measures. Current projects include measuring student growth in early elementary grades, defining chronic absenteeism, developing teacher and student surveys of school climate, and examining whether students are loved, challenged, and prepared.

As I observe the accountability work in the mid-Atlantic states and elsewhere in the country, I see two interesting directions for the future.

1. Accountability systems are broadening to include not only additional student outcomes but also school operations. This is partly a result of the ESSA provision that calls for an “additional measure of school quality or student success.” By far the most common “additional measure” is chronic absenteeism, now included in 37 ESSA plans. The REL has partnered with the state of Pennsylvania to understand the statistical properties of its chronic absenteeism measure and the implications of changing it.

Meanwhile, some states and districts are starting to measure students’ social-emotional learning (SEL), though they are exercising appropriate caution in attaching high stakes to these measures. The REL is working with the District of Columbia Public Schools to analyze its SEL data to produce an indicator of the proportion of students who are loved, challenged, and prepared. The district plans to use the data to better support students and promote districtwide attention to improving SEL.

Interestingly, some states have gone beyond student outcomes for their additional measures and are developing or implementing school climate surveys for accountability purposes. These typically involve surveys of school staff and students (and sometimes parents). For the first time, these surveys are incorporating information about what is happening inside schools into federally mandated accountability systems.

Information on school climate can be incredibly useful because student outcomes alone are not enough to help schools and districts figure out how to improve. Here at the REL, we’ve been working with Maryland to develop its student and staff surveys and to produce reliable indices measuring the climate of each public school in the state. I see this work as enormously promising because measures of school climate can complement measures of student outcomes by indicating the causes of these outcomes and suggesting directions for improvement.

2. Examining student progress over time to assess school performance can extend beyond reading and math in grades 3 to 8. Data that follow each student from year to year can be valuable in distinguishing a school’s impact—and thus its true performance—from the influence of factors that aren’t under a school’s control, such as family background. States have used these data to measure schoolwide student growth, but typically only in reading and math in grades 4 to 8. The longitudinal data systems that have been constructed over the past couple of decades are increasingly including student data that begin in kindergarten and extend beyond high school to college and the workforce. This is creating new opportunities to incorporate student progress or growth into school accountability measures:

  • Maryland now has a statewide kindergarten-entry assessment, and we are working with the Maryland State Department of Education to assess the possibility of producing a valid and precise school-level growth measure that follows students from kindergarten entry through grade 3. This work could be useful not only for accountability but also for diagnostic purposes by providing schools and districts with their first available evidence on the contributions of the first four grades of elementary schools to student growth.
  • Mathematica researchers are working (outside the REL) with Louisiana to measure the promotion power of each of the state’s high schools in enhancing students’ likelihood of graduating, enrolling, and persisting in college and, eventually, earning a living.

Other states might want to explore kindergarten entry assessments and longitudinal links to college and beyond to fill holes in their accountability systems and produce a picture of school performance that is more comprehensive, rigorous, and robust. Incorporating non-test outcomes such as graduation, college enrollment, and earnings could be particularly helpful to balance a focus on test scores. States could even include civic outcomes among their school performance measures, assessing the extent to which each high school increases the likelihood that its students eventually register and vote. Imagine evaluating schools based on their success in promoting civic education, the original purpose of public education!

Examining long-term student outcomes in a way that assesses the school’s effect on these outcomes is key to success in accountability. Using the raw outcomes (unadjusted for factors outside the school’s control) would recreate one of the flaws of No Child Left Behind: penalizing schools serving disadvantaged students and artificially inflating the performance of schools in wealthy communities. In the previous examples, both Maryland and Louisiana show that they understand the importance of applying sophisticated statistical approaches to their longitudinal data so they can measure schools’ impacts independent of factors outside the school’s control.

In the mid-Atlantic region and across the country, accountability systems are broadening and deepening. These changes should ultimately provide a richer understanding of school performance, fairer comparisons among schools, and better diagnostic information for leaders engaged in school improvement.