In education policy discussions, the word “accountability” has become synonymous with testing. However, its actual meaning has more to do with accepting responsibility for one’s actions — and a willingness to take action when things have gone wrong. So when it comes to public education, who actually is responsible?
At first glance it may seem the answer is “public schools” — that’s certainly the knee-jerk reaction when things are seen as having gone wrong. Although most educators lament that “we need to do better,” that answer is only partially correct. If one looks at achievement test scores (as a proxy for learning), research suggests only about one-third of differences in scores are attributable to what happens in schools. The other major factors in a child’s education are influences outside of school — home, community and broader educational policy, including testing, funding and how funding is distributed. This isn’t to suggest schools are irrelevant. This same research also consistently shows the single most important factor in a child’s education is his or her teacher.
The answer then to the question of who’s responsible and ultimately accountable for public education is really “we all are.” This notion is captured by the adage “it takes a village to raise a child.”
In recent weeks, accountability of the State of Texas for adequately funding public education has been on display with much attention to how, how much and for whom public education is funded. It is clear to anyone paying attention that the way public education is funded in Texas is broken and that the time has arrived for a major policy overhaul. As education scholars Wes Null and Diane Ravitch (Waco Tribune-Herald contributors, Jan. 26) pointed out, the last time Texas did a major overhaul of school funding was in the late 1940s. Other Trib contributors (Tarrant County Judge Glen Whitley, Feb. 2, and Greater Waco Chamber of Commerce policy director Jessica Attas, Jan. 5) confirm the clear constitutional responsibility the Legislature has for equitably funding public education. They point out ways the current system falls short — low overall funding for public education (41st nationally) and shifting predominant public education funding from the state to local property taxes (up from 52 percent to 61 percent in just the last 10 years).
It’s encouraging that proposals are now on the table for such things as local tax relief through greater state funding of schools as well as improved teacher salaries and support for badly needed professional support staff (e.g., school counselors), bolstering school security and acknowledging needs of schools with more challenging student populations.
Not addressed, however, is a re-examination of the state’s expensive and in some ways onerous system for monitoring student outcomes. The current system requires standardized achievement testing of all students in Texas public schools at each target grade level. Based upon test results overall and disaggregated for certain sub-groups, schools are assigned ratings that purport to reflect the quality of education provided by individual campuses. There are several major issues with this approach, not the least of which is the enormous cost financially and logistically.
First, current testing focuses primarily on academic knowledge outcomes of public education, which are only part of the important outcomes for a public education in a democratic society. Focusing only on these outcomes, while important, discounts (perhaps even dismisses) the importance of other critical outcomes such as responsible citizenship, tolerance of differences, time management, empathy, impulse control, etc.
Second, considerable research has shown that such standardized achievement testing results predominantly reflect the general socioeconomic status of the children attending a given school, not the quality of the educational experience being provided. Texas data from 2016, for example, shows that the vast majority of the 378 schools assessed by the state as Improvement Required were disproportionately high-poverty, high-mobility and majority-minority with higher percentages of second-language and special education students. So using standardized achievement tests as a high-stakes measure for gauging school effectiveness is flawed by measuring only one area of desired school outcomes. It distorts results by using a measure that conflates student academic outcomes with a school student’s socioeconomic makeup.
Third, by their very nature, standardized tests do not provide useful information about individual children’s academic knowledge. Such testing is designed primarily to reflect information about a group of test-takers; thus, it provides little useful information for guiding subsequent instruction for any given child.
Fourth, if the primary purpose of any statewide student testing is to monitor and inform state leaders and the public about the effectiveness of public schools based on student outcomes, even narrowly defined academic outcomes, there’s no reason to test every student in the targeted grade levels. Indeed, quality-control checks in industry and much of academic research successfully use a different paradigm. In those arenas, test results from randomly selected samples from a larger group are assessed with data compiled and analyzed for useful conclusions about relative quality or comparative differences that can then be generalized back to the larger group.
In educational testing, we often wish to know if particular educational treatments are differentially effective with different groups of students (e.g., schools attended, genders, ethnicities). The results of tests administered to randomly selected students from each identified subgroup of students (a process referred to as stratified random sampling) are compared to each other subgroup as well as to the combined groups. Statistically, these procedures provide a level of confidence (i.e., probability) that any observed score differences among sample groups are reasonably true (or not) and generalizable. In using such an alternate paradigm to monitor student outcomes, costs can be reduced without sacrificing accuracy.
A final major testing issue is whether an achievement standard or a growth standard should be used for making judgments about a school’s attainment of desired outcomes. A growth standard examines the progress a school is making toward a target student performance, whereas an achievement standard examines proximity (how close/whether) students collectively meet or exceed a target student performance. In schools where students widely underperform an achievement standard, there is a rather low probability for meeting that standard any time soon, especially given the many correlated factors which are beyond the school’s scope or resources to address. Given that these schools disproportionately serve high-poverty, high-mobility and majority-minority students, the prospect of a school fully meeting an achievement standard is daunting and demoralizing — for teachers, for students and for communities. What’s more, punitive consequences for schools whose students widely underperform an achievement standard rarely have much positive effect.
By contrast, a growth standard positively recognizes progress toward a target performance; for schools whose students are widely underperforming to show progress toward a target performance, the challenge is far less daunting, even if falling short of the target in the near term. While the goal for all schools remains student outcomes that meet or exceed target achievement standards, having a temporary/transitional period — where a growth standard allows low-performing schools to be successful by showing increasing progress toward the achievement standard — provides an incentive not devoid of hope or enthusiasm toward ultimately reaching the target performance.
While schools and the teaching profession have specific responsibilities and are accountable for the technical aspects of creating an educational environment that promotes high-quality student learning, all of us as citizens are accountable for having a strong system of public education. Each of us has our own unique responsibilities as parents, employers and related community agencies — and that includes state and local leaders. For its part, the Texas Legislature is constitutionally responsible for ensuring schools are adequately funded, setting macro-level educational policies and monitoring school performance on behalf of the people of Texas.
After a half-century hiatus, the Legislature is poised to address comprehensive funding reform and related policies. Not unrelated to this effort, for cost reasons, should be a re-examination of the state’s approach to monitoring school effectiveness — most importantly student outcomes. Previous efforts to monitor student learning in public schools have led to heavy dependence upon standardized achievement testing that is unnecessarily expensive, conflates student learning with students’ socio-economic status, measures only one area of student outcome (academic knowledge), is of little value in guiding instruction for individual children and is enacted in a high-stakes manner discouraging in academically low-performing schools where student performance is especially impacted by non-school factors. It’s time for the Legislature to acknowledge the limitations of monitoring only academic learning. Even if still confined to academic knowledge outcomes, lawmakers must revamp how those outcomes in public schools are monitored to include both achievement-based standards and transitional growth-based standards, especially for low-performing schools impacted by such problems as poverty. Just as simple solutions for complex problems are rarely effective, so is a simplistic approach to monitoring student learning and school effectiveness.
Adding nuance to the state’s monitoring responsibility by blending achievement and growth standards can help ensure schools and student learning are all moving in a positive direction and toward meeting or exceeding our collective expectations for students and schools — and at considerably less cost.