More than 100 education researchers in California have joined in a call for an end to high-stakes testing, saying that there is no “compelling” evidence to support the idea that the Common Core State Standards will improve the quality of education for children or close the achievement gap, and that Common Core assessments lack “validity, reliability and fairness.”
The California Alliance of Researchers for Equity in Education, a statewide collaborative of university-based education researchers, recently released a research brief (see in full below below) describing concerns with the Common Core standards and the assessments being given to millions of students in California and other states around the country this spring.
The researchers, from public and private universities in California — including Stanford University, UCLA, and the University of California Berkeley — say that the Common Core standards themselves do not accomplish what supporters said they would and that linking them to high-stakes tests actually harms students. The brief says:
Although proponents argue that the CCSS promotes critical thinking skills and student-centered learning (instead of rote learning), research demonstrates that imposed standards, when linked with high-stakes testing, not only deprofessionalizes teaching and narrows the curriculum, but in so doing, also reduces the quality of education and student learning, engagement, and success. The impact is also on student psychological well-being: Without an understanding that the scores have not been proven to be valid or fair for determining proficiency or college readiness, students and their parents are likely to internalize failing labels with corresponding beliefs about academic potential.
More specific to California: a recent study on the effects of high-stakes testing, in particular of the CA High School Exit Examination (CAHSEE), found no positive effects on student achievement and large negative effects on graduation rates. The authors estimated that graduation rates declined by 3.6 to 4.5 percentage points as a result of the state exit-exam policy, and also found that these negative effects were “concentrated among low-achieving students, minority students, and female students.”
The Common Core State Standards initiative has become a political issue, with Republican presidential candidates, including front-runner Donald Trump, repeatedly saying that if they become president, they will get rid of the Core. In fact, no president can do that with executive power. While the Obama administration supported the development of the Core and dangled federal dollars in front of states to “persuade” state legislatures to adopt the standards, 45 states and the District of Columbia each separately went ahead and approved the math and English standards (though some later decided to repeal or replace the standards). The administration provided two multi-state consortia with some $360 million in federal funds to develop new Core-aligned standards tests, which states could choose to join. The federal government can’t directly dictate to a state what standards and curriculum it must use.
Early in the development of the Core, there appeared to be little controversy and bipartisan support. But after implementation began several years — and was botched in many places — concern began to grow from across the political spectrum, for different reasons. Some educators and researchers questioned the way the standards were written (whether, for example, there was any or enough input from working teachers) and some criticized the content of the standards, especially for young children. Some critics said standards-based education has never been shown to be effective, and others felt the administration’s involvement usurped local authority. Tea party members and even the Republican National Committee jumped onto the anti-Core bandwagon, accusing the administration of a federal takeover of public education, extreme right-wing rhetoric that clouded a real discussion about the Core.
The administration’s support for the Core was one of the issues that propelled critics to accuse the U.S. Education Department of micromanaging local education issues and pushed Congress to finally move, last December, to pass a successor bill to the widely disliked K-12 No Child Left Behind law. The new law, the Every Student Succeeds Act, sends back from the federal government to the states a good deal of education policymaking power that had been used by the Obama administration in an unprecedented manner.
The brief also says:
Overall, there is not a compelling body of research supporting the notion that a nationwide set of curriculum standards, including those like the CCSS, will either raise the quality of education for all children or close the gap between different groups of children. Therefore attaching high-stakes testing to the CCSS cannot be the solution for improving student learning.
Yet, with the CCSS comes even more testing than before, and based on those test scores, any number of high-stakes decisions may follow, all of which are decisions using scientifically discredited methods, namely, the use of value-added modeling that purport to attribute gains in test scores to such factors.
California adopted the Common Core State Standards in 2010 and chose to align with one of two multi-state testing consortia funded by the administration. California signed up with the Smarter Balanced Assessment Consortium, which produced a testing system known as the SBAC. That group, as well as the Partnership for Assessment of Readiness for College and Careers (which produced the PARCC assessments), contracted with leading testing companies, including Pearson, the Educational Testing Service, and CTB/McGraw-Hill, to develop the new exams. In 2013, California began pilot-testing the SBAC exams, and in spring 2014 conducted widespread field testing. Later that year, high-stakes summative and interim standardized SBAC tests were administered; in spring 2015, several million students took the SBAC, and, as predicted a majority of students failed.
The brief says:
Testing experts have raised significant concerns about all (SBAC, PARCC, Pearson) assessments, including the lack of basic principles of sound science, such as construct validity, research-based cut scores, computer adaptability, inter-rater reliability, and most basic of all, independent verification of validity. Here in California, the SBAC assessments have been carefully examined by independent examiners of the test content who concluded that they lack validity, reliability, and fairness, and should not be administered, much less be considered a basis for high-stakes decision making. When asked for documentation of the validity of the CA tests, the CA Department of Education failed to make such documentation public. Even SBAC’s own contractor, Measured Progress, in 2012 gave several warnings, including against administering these tests on computers.
Nonetheless, CA has moved forward in full force. In spring 2015, 3.2 million students in California (grades 3-8 and 11) took the new, computerized Math and English Language Arts/Literacy CAASPP tests (California Assessment of Student Performance and Progress). The tests were developed by SBAC, and administered and scored by Educational Testing Service. Scores were released to the public in September 2015, and as many predicted, a majority of students failed (that is, were categorized to be below proficient). SBAC itself expected that pass rates would go down, and would be particularly low for certain groups, including English-language learners (who make up over 22% of the enrollment in CA public schools), whom SBAC predicted would see an approximately 90% fail rate.
The impact in California of high-stakes assessments (CCSS or otherwise) is not hard to predict. A compelling body of research exists on problems with high-stakes testing that range from the scientific discrediting of high-stakes testing to the disparate impact of high-stakes testing that further widens educational inequities.
The research brief also cites these concerns:
*The cost of implementing the CCSS assessments is high and unwarranted. The CCSS testing costs for CA are estimated at $360 million dollars in federal tax dollarsand $240 million dollars in state funds for three years of administration and scoring. The CA general fund appropriation for pupil testing in the 2014-2015 school year was $126,850,000. In practical terms, this means that standardized testing has taken precedence over other priorities such as class size reduction, quality teacher training and retention, programs in the arts, adequate science and technology equipment, and keeping neighborhood schools open.
*[T]he technology and materials needed for CCSS assessments require high and unwarranted costs. Much of these additional costs relate to the computer-based assessments, which require upgrading equipment (computers, headphones, keyboards), bandwidth (for data-heavy tests that include videos, animated graphics, and interactive charts), and technical support in a short period of time, which means that already-struggling schools will be disproportionately impacted.
*The technology requirements raise concerns not only about cost, but also about access. The CCSS assessments involve computer use not only for the actual assessments, but also for the practice assessments, and both require that students have connectivity, computer access, and computer familiarity. As such, CCSS assessments favor middle- and high-income students who typically have easier access to technology, Internet connectivity, and keyboard practice both inside and outside of school.
*[T]he CCSS assessments have not provided for adequate accommodations for students with disabilities and English-language learners, or for adequate communication about such accommodations to teachers.
The researchers include a list of recommendations, including taking a new look at appropriate standards and assessments, and placing a ban on high-stakes testing until specific questions about proper accountability for schools and students and teachers have been addressed.
Here’s the full research brief: