Future Standardized Exams Will Focus on Informing Instruction
If you or someone you know has taken the SAT, or an Advanced Placement test, or even so little as considered college in the past 12 years, chances are APS Charter Member Howard Everson has affected your life, and you didn’t even realize it.
That’s because Everson is the chief research scientist at the College Board, where he has worked in some research and development capacity since 1992. He is among the many scientists at the College Board who develop and create the SAT, PSAT, AP, and NMSQT tests awaiting high school juniors and seniors across the country each year.
Everson’s roots in educational measurement extend all the way back to his graduate days, when his interest in human learning – particularly learning from schooling – served as the impetus for his doctoral study in educational psychology, as well as his postdoc work in psychometrics at the Educational Testing Service.
“I became convinced that I had to learn about the role of measurement, to better design interventions and measure learning outcomes,” Everson said.
That sequence of studies exalted Everson into the pantheon of American test makers, and perhaps the nightmares of teenagers across the United States. In 2003, a little over 1.5 million students took the SAT or SAT II subject test, the apogee of a number that has grown steadily for years. But regardless of how many young men and women take standardized tests, the measurement principles behind their creation remain basically the same.
“Those of us working in the field of educational measurement and testing are attempting to create high quality psychometric measure,” he said. “We’re working to create tests and assessments that meet the standards of validity, reliability, and fairness as set forth by professional psychologists and educators in the United States.”
In order to reach this lofty and expansive goal, Everson and his colleagues rely first and foremost on a complete understanding of the constructs of intelligence, motivation, and achievement. They work to design assessments that are malleable enough to measure these psychological constructs while remaining robust to extraneous factors, such as varying administration conditions, and other unpredictable events. The results are “standardized” conditions of testing – time, language, length, content, and format – essential to accurate results.
“Variation in those extraneous aspects of a testing program can affect the test scores and the inferences drawn from those scores, just as contaminated lab procedures affect the results of a blood test, an X-ray, or a biopsy.”
The comparison to other sciences is completely in line with Everson’s approach toward test design and development, which combines his background in measurement with theory on how cognitive abilities develop. “This effort has pushed me to explore ways in which cognitive science and psychometrics could be joined and integrated,” he said. “Coupled with sound theory, good measurement principles can and do yield high quality, useful standardized tests.”
According to Everson, traditional forms of standardized tests are often criticized for losing sight of the underlying psychological constructs and processes that drive performance and achievement. This may explain why most of his biggest influences are psychologists like Edmund Gordon, Robert Sternberg, Robert Mislevy, Sigmund Tobias and even William James, whom he quotes on the College Board Web site on the importance of teachers to America’s future.
“Our biggest challenge, as a field, is to move forward with a more complete understanding of learning, instruction, and cognition within the context of specific academic domains, such as mathematics, critical reading, science, writing, etc. We also need to focus more on the cognitive processes that are necessary for success on academic tasks,” he said.
But understanding learning and cognition are not even the College Board’s most pressing challenges. Despite their proven effectiveness, standardized assessments remain heavily maligned and scrutinized across many public and political arenas. The harshest critics contend that the tests are “culturally biased” and fail to embrace the egalitarian spirit under which they were conceived. Everson called such a view “misbegotten and unfortunate.”
“Our enduring critics are those who believe that standardized tests are anathema to education and student learning,” he said. “But tests like the SAT and AP have done a great deal to promote the twin goals of academic excellence and equity. The evidence is seen in the growth in these testing programs and the increased access to higher education over the past 50 years.”
In the summer of 1999, however, the College Board’s testing programs received a sudden growth spurt, when former West Virginia governor Gaston Caperton assumed the Board’s presidency. “Caperton brought a vision that animates much of the Board’s work today,” Everson said. “He talked of not simply documenting students’ achievements with our testing programs, but of shifting our focus to promote student achievement and increase the academic preparation of students.”
The movement to not only test college preparation but rather to enhance it has permeated all branches of the College Board and now prepares to burst though the sluice gate into mainstream American culture. “I see a move toward more achievement based testing with more of an emphasis on content and performance standards. I also see the need for more tests that can inform instruction – tests that can tell teachers what students know and do not know about a subject,” Everson said.
In keeping with the field’s current trends, Everson’s research interests extend beyond cognition, instruction, and assessment into the less cultivated landscape of designing tests that inform instruction. “I refer to these [tests] as instructionally sensitive assessments,” he said. “I’m also interested of late in issues of instructional design that promotes student learning and works to close the achievement gap between minority and non-minority students in the United States.”
In fact, about the only constant in the changing world of test creation is the desire to get a leg-up on the competition. And since having the test creator answer your questions is a table-turn the likes of which might occur once in a lifetime, the following question was too tempting to resist: Is there any truth to the myth that a test-taker should, when all else fails, choose letter “C?” To the overwhelming financial relief of testing agencies, and to the discomfiture of pre-matriculating students and parents across the country, the response was an unhesitating and scientifically resonant, “No.”
Leave a comment below and continue the conversation.