At a time when more high schools are looking to their graduates’ college-remediation rates as a clue to how well they prepare students for college and careers, new research findings suggest a significant portion of students who test into remedial classes don’t actually need them.
Separate studies from Teachers College, Columbia University, and the Harvard Graduate School of Education come to the same conclusion: The way colleges are using standardized placement tests such as the College Board’s Accuplacer, ACT’s Compass, and others can misidentify students, and secondary schools and universities should work to develop a more comprehensive profile of students’ strengths and weaknesses in performing college-level work.
The problem is coming to the fore as more states move to align their academic standards for college and career readiness with the Common Core State Standards and federal Race to the Top requirements and more high schools receive data on how their graduates are faring in colleges.
Thomas W. Brock, the new commissioner of the National Center for Education Research and a veteran higher education researcher, said improving remedial education has become a top research and policy concern. “It’s a huge need,” he said. “At many institutions, it’s a majority of students coming in and being placed into developmental ed.—and this is where it starts to bleed into the financial-aid agenda, because they’re using up valuable semesters of financial aid, which of course are not endless.”
Growing Issue
According to the most recent data from the National Center on Education Statistics, 3 million new students enter higher education each year, and fully half take at least one catch-up course while they are enrolled—at a cost of nearly $7 billion a year for the noncredit-bearing classes.
“If you’re working in community colleges, especially urban community colleges, you get used to those numbers,” said Judith Scott-Clayton, an assistant professor of economics and education at Teachers College and an author of one study, which was published under the auspices of the National Bureau of Economic Research and presented at last month’s American Economics Association meeting in San Diego. “Remediation is the typical experience now.”
Those high rates of remediation have long been used by education policymakers to suggest that primary and secondary schools do not prepare students adequately for college-level work. They were one of the key arguments behind the development of the common core and other standards-reform initiatives, and states such as Illinois include remediation rates in feedback reports to high schools.
“It’s being used in some places for high school accountability, so this certainly raises a word of caution,” said Ms. Scott-Clayton. “We can’t just take the remediation rate as purely objective and without problems. Is this accountability gone awry?”
Ms. Scott-Clayton analyzed the high school transcript information, college-placement-test scores, and college progress of more than 50,000 students in a large, unnamed urban community college system and a separate statewide college system in the same state, which also is not identified. In the urban college system, which includes a high-minority population, she found three out of every four students were assigned to at least one remedial class; nearly 80 percent of all those who took the mathematics placement test were assigned to a remedial course. In the statewide system, the numbers were only slightly better: 70 percent of students who took the math placement test and 58 percent of those who took the English test were assigned to noncredit-bearing catch-up classes.
To determine whether all those students were really so unprepared for college-level work, Ms. Scott-Clayton examined the students’ actual high school and college credits earned and grades received. She found that 20 percent of students placed in remedial math and 25 percent of those placed in remedial reading were “severely misidentified,” meaning that not only could they have passed the entry college course in that subject, but they could have done so with a grade of B or better.
“So much focus and attention is on the students who are not succeeding, and how to help them succeed, that there just was not much talk about, ‘Hey, maybe there are some people in remediation who don’t need to be there,’” Ms. Scott-Clayton said.
In response to the placement-test studies, “What I will definitely say is the results we’re seeing are not surprising at all,” said David Parmele, the executive director of the College Board’s Accuplacer in Reston, Va. “Remedial education in America is broken right now, and we think there is a lot more that can be done to help.”
Mr. Parmele and representatives of ACT Inc., the Iowa City, Iowa-based nonprofit that runs the Compass placement test, said they agreed with the research methodology. Still, Jim Sconing, an ACT assistant vice president for applied research, cautioned that not all students enter college with high school transcripts or other indicators to weigh into their class placement. In particular, students who were homeschooled or who return to college after several years in the workforce may have no other information available. According to the most recent data from the National Center on Education Statistics, nearly a third of all entering college students are not coming directly from high school.
Mr. Parmele argued that Accuplacer does provide appropriate information on placements, but that colleges are more likely to misidentify students who do not prepare for the test, particularly if the placement test is used without any other academic indicators. Mr. Parmele and Mr. Sconing said both Accuplaccer and Compass include tools to allow colleges to weigh test results along with other academic indicators, such as high school grades and course credits, and work with colleges to use broader measures of student readiness than just the test. Neither testing representative, however, knew how many of its client colleges actually use those tools.
Broader View
Broader measures do increase accuracy, Ms. Scott-Clayton said; she found 30 percent lower rates of severe misidentification when placement decisions incorporated high school transcript data, including a student’s GPA, total courses taken and credits earned, honors classes taken, total classes in English and math, and the number of F grades received.
“The high school transcript info is basically more accurate for every group we look at,” Ms. Scott-Clayton said. “It’s true that it’s more subjective, but you are getting multiple measurements accumulated over time across several instructors. And it is capturing a broader array of skills, not just pure mechanical test-taking skills, but effort, persistence, motivation—things that we know matter a lot for college success.”
However, she found results uneven for specific racial groups: While fewer Hispanic students were placed in remedial courses using high school information, more young black men were pulled into remediation.
As part of researchers Angela Boatman and Bridget T. Long’s developmental-placement project at Harvard’s Strategic Data Project, Lindsay Daugherty, an associate policy researcher with the RAND Corp., also found test-based misplacement in a large school district and community college system in Texas. They found that at several community colleges requiring universal placement tests, 30 percent of students assigned to remediation were considered “college ready” based on their scores on the ACT, SAT, and state Texas Assessment of Knowledge and Skills, or TAKS, readiness assessments.
“Everyone understands what a good SAT score or ACT score is. It’s much more difficult, even within a local district to find the cutoff scores,” Ms. Daugherty said.
Aaron J. Graczyk, a senior project manager at the Texas Higher Education Coordinating Board, agreed. He said the state found commercial placement tests “were not as well-aligned as we would have liked” with Texas’ college-readiness standards, which span kindergarten through college.
New Rules in Texas
In light of that, and of high remediation placement, particularly in community colleges, the state legislature has required all colleges in the state to move to a new, state-developed placement test, which will roll out this fall. The new assessment includes diagnostic tests to identify specific problem areas in each subject, and the coordinating board will require colleges and universities to use those to plan more targeted remediation—for example, enrolling a student in a credit-bearing class while providing tutoring or an elective class to fill gaps in the student’s knowledge in that subject.
“If you are placed in remedial education,” said Judith Loredo, the assistant commissioner for the P-16 initiative of the coordinating board, “the diagnostic part of the test will help the institution better provide the services you need based on your weaknesses, and not on your strengths, so that you don’t end up necessarily having to take a 15-week course, go into this sinking hole, and rot for the next two years.”
It’s easy for students to end up in that “sinking hole,” Ms. Daugherty and Mr. Parmele agreed, because students often are less aware of the importance of the placement tests before they take them. While academic counselors and college recruiters teach students the importance of getting a good score on a college-entrance exam, and prep courses abound, there is little advance warning or preparation for placement tests.
“Many students are given placement tests without even realizing they’re going to be taking a test” Mr. Parmele said. “It’s hard for anybody to sit down for a test with no notice and try to remember what you learned about dividing polynomials in a class four years ago.”
Ms. Daugherty found the Texas students often did not know how the tests worked, what a remedial class was, or “that they wouldn’t get college credit but they would have to pay for it and it could be up to a year of remedial courses,” she said. “These tend to be higher-performing, low-income minority students whose parents don’t have the time and resources to worry about whether a placement test affects their kids.
“Students don’t realize what they’ve gotten themselves into until they’re halfway into the course,” she said.
The Strategic Data Project has been working with Texas districts to encourage more high school educators to talk to their students about placement as well as entrance exams, and to give exiting seniors a brush-up course before they take the placement tests. In Ms. Daugherty’s study, to be presented at the March meeting of the Association for Education Finance and Policy, she found those refresher courses can reduce remediation placements.
“Most of these higher math-performing students hadn’t been in algebra for three to four years, maybe,” she said. “They needed to just sit down for three or four hours and do some algebra problems, and it would come right back to them.”
