Test-based accountability: The best we can do?

By Sandra Stotsky | October 19, 2016, 11:27 EDT

Who doesn’t want accountability in education? That was the selling point for No Child Left Behind (NCLB). It was the selling point for the Massachusetts Education Reform Act (MERA) of 1993. It was even a selling point for Finnish reforms in teacher education in 1970.

But how did Congress and most of our education researchers and policy makers get frozen into thinking that test-based accountability is the only kind of accountability?

There are other forms of accountability—and maybe, just maybe, another form of accountability, or even hybrids of existing forms, might lead to gains in U.S. student achievement, if indeed that is a goal of accountability (i.e., to ensure that federal taxpayers are getting something worthwhile for their money).

Improved student achievement was the goal of MERA in 1993. Schools and school districts were held accountable. (The “grand bargain” was money for more effective schools.) Over the next decade, many changes took place in public education, including increased funding, “revised learning standards,” “revised student assessments based on clear curriculum frameworks,” and “revised teacher licensing.”

These reforms eventually did lead to increased levels of achievement for all groups of kids, even though average differences in achievement between groups were not closed after 10 to 15 years. (Closing achievement gaps was not a goal of MERA; improving academic achievement for all students was.)

Gains in student achievement were a goal of Finnish reforms in 1970. Professional accountability is how Helen Ladd describes the Finnish version of accountability. Schooling was restructured (enrollment in a free upper secondary school—vocational or academic—was encouraged, even though students could legally leave school after grade 9); teachers’ academic quality and teacher education were academically upgraded; and teachers in grades 1 to 9 were in charge of their school’s curriculum, with minimum guidance from a central ministry of education and no state-mandated testing in grades 1 to 9. See this review for details.

And gains it did lead to, for large numbers of children—not for all, and not without some problem areas, but, overall, good enough that Finland has not chosen to undo its reforms in the structure of schooling and in admission requirements for teacher education programs, so far as we know.

On the national level, No Child Left Behind, passed by Congress in 2001, also sought to increase student achievement. But unlike the Massachusetts reforms, NCLB did not lead to substantive gains in any of the three areas tested or to academically stronger teachers overall.

Yet NCLB seems to have fortified the erroneous belief that test-based accountability was the only way to strengthen the education of under-achieving or low-performing students despite the lack of long-term evidence for this policy. Nor did education researchers themselves come up with alternatives to the extensive testing regime mandated by NCLB (every grade from 3 to 8 and once in high school in both mathematics and English language arts, plus three science tests at different grade levels) and reauthorized in the 2015 Every Student Succeeds Act (ESSA).

Congress didn’t seem to care that many years of test-based accountability had not actually improved the education of low-performing students, to judge by NAEP test results. Nor did Congress seem to care that test-based accountability had had a deleterious effect on the school curriculum, on teacher behavior, and on student achievement. Because, unlike MERA and NCLB, ESSA was never about improving academic performance for all students.

But it was clear to observers that all was far from well in the public schools themselves. Opt-outs (students whose parents had kept them from taking federal- or state-mandated tests) and home schooling increased from 2015 to 2016, state participation in one of the two U.S. Department of Education-funded testing consortia (PARCC and SBAC) dwindled to a small minority of states, and school administrators were beginning to call public attention to the damage caused by test-driven accountability system being implemented by education bureaucrats.

Are there no other possibilities for an accountability system than this? Is annual testing and a teach-to-the-test classroom environment the only outcome possible in the search for a holy grail—a way to hold someone accountable for group differences in student achievement if student and parent responsibility was to be held off-limits?

Interestingly, no attention seems to have been paid to Helen Ladd’s articulation of alternative accountability systems in 2012. Now a Senior Fellow at the Brookings Institution, she continues to call for attention to inspection-based accountability systems, used by many of the countries whose students perform better than ours do on international tests of mathematics and science knowledge (Trends in International Mathematics and Science Study—TIMSS). She even wants experiments with such a system, as if Congress were really interested in more than gap-closing, something the USED will soon manage from behind its green curtain as test scores roll in from working computers in every school system in the near future and the USED can declare without opposition who has passed.

The current test-based accountability systems do not work and do not seem fixable.

It’s time for fresh ideas on accountability. It is not that we should give up on tests altogether. But no country with a functioning public education system hinges the flow of taxpayer funds on student scores from a barrage of tests, especially when it is unclear whether the tests are even academically sound.