States should not use the SAT or ACT to measure high school achievement because those exams don鈥檛 fully reflect states鈥� academic standards, and could distort what鈥檚 taught in the classroom, according to a study released Tuesday.
The paper, released by Achieve, which pushes for high-quality standards and tests, calls for a halt in an assessment trend that鈥檚 been picking up steam in recent years: states using the SAT or ACT instead of their high school tests. This school year, 13 states are using one of those college-admissions tests statewide to measure high school achievement.
Testing experts have long raised questions about using the SAT or ACT as an official measure of high school achievement. They argue that college-admissions exams can鈥檛 do a good job of measuring students鈥� mastery of their state鈥檚 academic standards because they were not crafted to gauge students鈥� grasp of those standards. They were designed to do something different, which is to predict students鈥� chance of success in college.
At the heart of Achieve鈥檚 paper is this question: How well do the SAT and ACT measure mastery of a state鈥檚 academic standards?
In the world of testing, that鈥檚 called 鈥渁lignment,鈥� and it鈥檚 a core principle of the field that a test has to be well aligned to what it鈥檚 trying to measure鈥攊n this case, the knowledge and skills in state standards鈥攊n order to produce valid results.
鈥榃e Respectfully Disagree鈥�
Both the College Board and the ACT disputed the new paper鈥檚 contention that their tests aren鈥檛 well enough aligned to states鈥� standards.
鈥淲hile ACT values the opinions of other experts in the field, we respectfully disagree with the contention that assessment of state academic standards and assessment of college preparedness are mutually exclusive,鈥� the company says in a statement.
The College Board said it has conducted studies showing that the SAT is aligned 鈥渨ith the current standards in all 50 states.鈥� The exam also 鈥渕eets or exceeds鈥� all requirements for state assessments in federal law, including that they align to state standards, the group said.
鈥淭he SAT is the only college-entrance exam that has been redesigned in accordance with the current state standards that guide teachers鈥� work in classrooms around the country,鈥� Michele McNeil, the College Board鈥檚 executive director of K-12 and global policy and external relations, says in a statement.
Achieve, which played a key role in developing the PARCC common-core test, examined three recent alignment studies. All three find the college-admissions exams lacking, although to differing degrees. Here are the highlights:
鈥� Study of the ACT鈥檚 alignment to the Common Core State Standards, by Achieve, 2018: Less than half of the questions in English/language arts, and less than half of the items in math, were judged to be aligned to the high school expectations in the common core. Questions that purported to measure mastery of writing standards didn鈥檛 ask students to produce writing. Achieve says that in its own technical manual, ACT acknowledges that 40 percent to 43 percent of its math items measure pre-8th-grade math content.
鈥� Study of the feasibility of letting districts swap Florida state tests for the ACT or SAT, by the Assessment Solutions Group, 2018: Neither the SAT nor the ACT fully cover Florida鈥檚 academic standards. Either test would require the state to add more questions to the college-admissions exams for either one to reflect Florida鈥檚 math and English/language arts standards. It suggests the revision or replacement of five to seven questions in the SAT, and 10 or more in the ACT. The study also found that results from the ACT and SAT would not be comparable to those from Florida鈥檚 own Algebra 1 or 10th grade English tests, 鈥渃asting doubt on the interchangeability鈥� of the three tests.
鈥� Study of the SAT鈥檚 alignment to the Common Core State Standards, by the Human Resources Research Organization, 2016, for Delaware and Maine: The report found that 76 percent of the English/language arts questions鈥攁 level it describes as 鈥渞easonably aligned鈥濃�攁nd 47 percent of the math items were fully aligned to the common core, which both states use as their academic standards. The report singled out particularly weak coverage of geometry and statistics, and recommended that states add more math questions to adequately reflect those standards.
Another alignment study, which was not included in Achieve鈥檚 report, found a 鈥渟trong or moderate match鈥� between the SAT and Connecticut鈥檚 common-core standards. That 2016 study, by a team from the University of Connecticut, found the weakest coverage of math standards in functions and geometry.
A similar study done for Wisconsin on the ACT, and also not included in Achieve鈥檚 paper, found that the ACT measures the skills and knowledge in the state鈥檚 academic standards. It notes, however, that in math, some ACT questions focused on skills 鈥渢ypically addressed at lower grade levels.鈥�
Alignment studies are wonky. But Achieve argues that they鈥檙e important because what鈥檚 measured on tests exerts a tremendous influence on what鈥檚 taught in classrooms.
鈥淭he danger in using admissions tests as accountability tests for high school is that many high school teachers will be driven to devote scarce course time to middle school topics, water down the high school content they are supposed to teach in mathematics, or too narrowly focus on a limited range of skills in [English/language arts],鈥� the paper said.
Achieve urges states not to adopt the SAT or ACT to measure high school achievement for accountability. Administering a college-admissions test statewide to get more students thinking about鈥攁nd applying to鈥攃ollege is an 鈥渁ppropriate鈥� use of those tests, but using them to measure achievement is 鈥渋ll-advised,鈥� the paper says.
A Good Match?
It鈥檚 a tricky policy issue, because when states opt for college-admissions tests instead of standards-based tests, they get benefits that resonate in the wake of a major backlash against overtesting, said Robert A. Schaeffer, the public education director of FairTest, which opposes high-stakes standardized tests.
鈥淭his is a two-edged sword,鈥� he said. 鈥淸Using the SAT or ACT] can reduce the volume of testing. And it saves parents money by shifting the cost of those tests from their checkbooks to the taxpayers.鈥�
But there鈥檚 still the question of whether the tests are a mismatch for measuring achievement, rather than college readiness, he said.
鈥淚t鈥檚 a truth-in-advertising thing,鈥� Schaeffer said. 鈥淚f you want a test to give you accurate information about students, teachers, schools or systems, you have to know that the test measures what it claims to measure.鈥�
In its paper, Achieve urges states not to use the SAT or ACT as statewide achievement measures. But it also urges caution if states are considering giving individual districts the freedom to substitute a college-entrance exam for a state high school test. That鈥檚 a new鈥攂ut still not-too-popular鈥攆orm of testing flexibility offered in the Every Student Succeeds Act. Allowing such a choice could create 鈥渟ignificant comparability issues鈥� between the college-admissions exams and state high school tests, Achieve argues. It might also let districts 鈥渟hop around鈥� for the test that puts them in the best light.
States that already use the SAT or ACT to measure high school achievement statewide should consider augmenting those tests with additional or revised items to get a fuller representation of their academic standards, the Achieve paper argues. That鈥檚 not unprecedented. Maine and Illinois, for instance, have augmented those exams to more fully cover their standards.
ACT spokesman Ed Colby said that the company is 鈥渁bsolutely committed鈥� to helping states if they want to augment the ACT, but maintains that augmenting 鈥渙nly supplements, rather than establishes, the validity of the ACT for ESSA accountability purposes.鈥�
The College Board did not respond to a question about its willingness to work with states that wish to augment the test.
