Opinion
Federal Opinion

Let’s Mend, Not End, Educational Testing

By Madhabi Chatterji — March 11, 2014 6 min read

The Common Core State Standards and accompanying K-12 assessments have recently sparked a fierce national backlash against testing. Sound educational testing and assessment are integral to good teaching and learning in classrooms and necessary for evaluating school performance and assuring quality in education. Rather than throw the baby out with the bathwater, I propose a more considered, “mend, not end” approach to testing, assessment, and accountability in America’s schools, with validity at the forefront of the conversation.

Mending begins with understanding that most commercial standardized tests are designed to serve particular purposes well, for particular populations, and can support only particular decisions at best. To uphold validity principles in practice, it is worthwhile to ask: Are we using the test for the originally intended purpose, or for another purpose that taxes the tool beyond its technical limits? Multi-purposing a test indiscriminately is not a good idea from a validity standpoint, despite its efficiency.

Validity deals with the meaningfulness of test scores and reports. Technically, validity is determined by the built-in features of a test, including its overall content, the quality of test questions, the suitability of metrics for the domains tested, and the reliability of scores. In addition, how and where a test’s results are applied, and the defensibility of inferences drawn, or actions taken, with test-based information affect the levels of validity we can claim from the scores and reports.

BRIC ARCHIVE

According to testing standards published by the American Educational Research Association, the National Council on Measurement in Education, and the American Psychological Association, once a validated test is taken out of its originally intended context, we may no longer be able to claim as much validity for a new population, purpose, or decisionmaking context, nor with as much certainty.

New proposed uses call for more tests of a test—a process called “validation.” New evidence must be secured to support a new or different action. Too often, this basic guideline is overlooked, particularly under high-stakes accountability policies like the federal No Child Left Behind Act or the common core. Validity oversights also happen with relatively low-stakes international-assessment programs like the Program for International Student Assessment, or PISA.

No Child Left Behind, signed into law in 2002, mandated testing of all students in grades 3-8 to measure progress of schools based on results of annually administered achievement tests. Variable state-set standards toward manifestly unattainable growth targets of “adequate yearly progress” and “universal proficiency” by 2014 stretched many school evaluation systems beyond their technical capabilities. NCLB’s public rewards and sanctions based on school performance led to “teaching to the test,” spuriously raising student test scores without lasting or replicable learning gains. This repercussion, in and of itself, undermined the validity of inferences from test scores, which no longer indicated clearly what students actually knew on tested domains.

Ripple effects of NCLB took hold in other school evaluation contexts, too, threatening validity in additional ways. Even the most enlightened and progressive of districts were pressured into missteps by high-stakes-testing requirements. In 2005, for example, Montgomery County, Md., sought to ratchet up performance and close achievement gaps districtwide by identifying its own model schools and school practices—a laudable goal. However, the county’s selected measure of student achievement, aggregated to serve as an indicator of school performance in “value added” evaluation models, was the combined math and verbal SAT score of high school students.

Recent efforts have sought to align the SAT more with college-readiness and common-core standards, but at the time of the 2005 report, “Value-Added Models in Education: Theory and Applications,” the validity of the SAT as an indicator of school-level outcomes was questionable. A college-entrance exam, the SAT is designed to predict how well students will perform as college freshmen, with limited validity as a curriculum-based achievement test. Variability in the levels and kinds of coursework taken by students could significantly affect the meaning of the scores, weakening inferences about student achievement in K-12 scholastic programs.

See Also

Check edweek.org Monday, March 17, for the launch of a time-limited group blog facilitated by Madhabi Chatterji and James Harvey. Assessing the Assessments will focus on measurement, assessment, and accountability.

Further, because students opt to take the SAT, test-takers are likelier to be stronger academically and inclined toward college, come from wealthier families, or have exposure to stronger schooling experiences. Self-selection biases schools’ aggregate SAT scores, complicating interpretations of what caused them to rise or fall.

Neither the school district nor the SAT is at fault. Rather, punitive accountability measures tied to test results in the larger context of reforms may be called into question. The power of such accountability mandates influences decisions of even trained analysts, regardless of stakes tied to local actions.

In the current context of the common core, a parallel drama is playing out. The common-core tests now being developed have been criticized as too long, superficial or overly narrow, and out of alignment with the curriculum and common-core standards. Educators, parents, and local officials reasonably fear that, yet again, tests are serving as blunt policy instruments to drive top-down reforms with inadequate time and resources for designing deeper curriculum and assessments to match, with little or no professional development of teachers and school leaders and in neglect of critical supports that schools need to succeed.

With ill-prepared schools and students, what will the test results really tell us about student learning and the quality of schooling?

Yet, were the same tests implemented after standards were refined, teachers and schools readied, parents and students oriented, tests validated to measure what students actually learned better, and results freed from external rewards and sanctions, the results might be more meaningful. Further, the anti-testing backlash might well disappear.

No one was celebrating the recently released results on the 2012 PISA, ranking American 15-year-olds below their peers in many other industrialized countries, particularly in math and science. But how meaningful and defensible are the intercountry comparative averages, given the differences in culture, educational opportunity, and backgrounds of 15-year-olds tested from different nations?

Despite popular claims, these sample survey statistics also cannot tell us much about whether particular regional reforms failed or succeeded. Interpreted carefully, PISA results yield useful benchmarks within particular nations, opening opportunities for education systems to improve.

Misinterpretation of PISA’s intercountry rankings, however, reflects a larger syndrome of misuse of educational assessment results and hand-wringing about public education that could easily be avoided.

Most standardized instruments rest on a solid base of scientific knowledge that dates back to the first half of the 20th century. These tools have documented achievement gaps in ethnic, gender, and socioeconomic groups reliably, furnishing policymakers, educators, and our society at large with evidence for improving conditions.

But misuse and misinterpretation of standardized-test results is a pervasive problem in educational assessment that threatens levels of validity, especially in high-stakes testing contexts. Here’s an area where scholars and practitioners; test-makers and test users; educators, parents, and students; and the media could work together to make a difference.

These and other issues will be open for debate and discussion in a time-limited blog hosted by edweek.org, to be launched next week and facilitated by James Harvey of the National Superintendents Roundtable and me. Assessing the Assessments: K-12 Measurement and Accountability in the 21st Century will feature expert commentary from scholars and practitioners, offering a variety of perspectives on today’s critical assessment challenges.

Related Tags:

A version of this article appeared in the March 12, 2014 edition of Education Week as Validity Counts: Let’s Mend, Not End, Educational Testing

Events

This content is provided by our sponsor. It is not written by and does not necessarily reflect the views of Education Week's editorial staff.
Sponsor
Equity & Diversity Webinar
Culturally Relevant Pedagogy to Advance Educational Equity
Schools are welcoming students back into buildings for full-time in-person instruction in a few short weeks and now is the perfect time to take a hard look at both our practices and systems to build
Content provided by PowerMyLearning
Classroom Technology Webinar Making Big Technology Decisions: Advice for District Leaders, Principals, and Teachers
Educators at all levels make decisions that can have a huge impact on students. That’s especially true when it comes to the use of technology, which was activated like never before to help students learn
Professional Development Webinar Expand Digital Learning by Expanding Teacher Training
This discussion will examine how things have changed and offer guidance on smart, cost-effective ways to expand digital learning efforts and train teachers to maximize the use of new technologies for learning.

EdWeek Top School Jobs

Teacher Jobs
Search over ten thousand teaching jobs nationwide — elementary, middle, high school and more.
View Jobs
Principal Jobs
Find hundreds of jobs for principals, assistant principals, and other school leadership roles.
View Jobs
Administrator Jobs
Over a thousand district-level jobs: superintendents, directors, more.
View Jobs
Support Staff Jobs
Search thousands of jobs, from paraprofessionals to counselors and more.
View Jobs

Read Next

Federal Biden Calls on Schools to Host COVID-19 Vaccination Clinics for Kids 12 and Up
The president is focusing on vaccinating children ages 12 and older as concerns grow about the Delta variant and its impact on schools.
2 min read
President Joe Biden speaks in the South Court Auditorium on the White House campus, Wednesday, June 2, 2021, in Washington.
President Joe Biden speaks in the South Court Auditorium on the White House campus on June 2.
Evan Vucci/AP
Federal How Political Backlash to Critical Race Theory Reached School Reopening Guidance
A lawmaker wants Miguel Cardona to repudiate the Abolitionist Teaching Network after federal COVID-19 documents referenced the group's work.
6 min read
Rep. Virginia Foxx, R-N.C., is seen at a press conference at the U.S. Capitol on March 9, 2021 in Washington.
Rep. Virginia Foxx, R-N.C., is seen at a press conference at the U.S. Capitol on March 9, 2021 in Washington.<br/>
Graeme Sloan/SIPA USA via AP
Federal Biden Team: Schools Can Go Beyond Trump Rules in Response to Alleged Sexual Misconduct
The Education Department's guidance, released July 20, states that Title IX rules from 2020 lay out "minimum steps" for educators.
3 min read
Symbols of gender.
iStock/Getty
Federal Fact Check: After Furor Over 1619 Project, Feds Adjust History and Civics Grant Plans
A previously obscure history and civics program has weathered a political storm, but what exactly has changed?
4 min read
Education secretary nominee Miguel Cardona speaks during a Senate Health, Education, Labor and Pensions Committee hearing on his nomination on Feb. 3, 2021, in Washington.
Education secretary nominee Miguel Cardona speaks during a Senate Health, Education, Labor and Pensions Committee hearing on his nomination on Feb. 3, 2021, in Washington.
Anna Moneymaker/The New York Times via AP