Holly Yettick is a doctoral student in the Educational Foundations, Policy and Practice program at the School of Education at the University of Colorado in Boulder.
College remediation rates are the school accountability measure du jour. Once relegated to the dusty realms higher education, a topic largely ignored unless it involves athletics or scandals (or athletic scandals), remediation rates took center court when Colorado and other states linked K-12 and postsecondary databases.
All of a sudden, we could evaluate individual districts and schools by examining what percentages of their graduates were assigned to remedial college courses.
Like too many educational measures, college remediation rates are often misused and misunderstood. Specifically, I have noticed that they are often mentioned in the same breath as high school graduation rates. By this I mean that they are treated as if they are alternatives to standardized exams—i.e. as holistic assessments of the ultimate outcome of 12-plus years of education.
They are not.
At least not in Colorado and many other states. According to Colorado’s remedial education policy, students are referred to remedial courses based upon their math, reading and/ or writing scores on ACT, SAT or ACCUPLACER exams. For instance, first-time undergraduates are referred to math remediation if they earn less than a 19 on the ACT math section, less than a 470 on the SAT math section or less than 85 on the ACCUPLACER Elementary Algebra test.
So unless you believe that these tests are holistic indicators of a student’s past performance and future potential (and some do), remediation rates are not alternative measures that prove or disprove the success or failure of the standardized testing movement or add a new, non-test-related dimension to the test score data already supplied by the state. At least not in Colorado. In Colorado, remediation rates ARE standardized exams.
There are certainly advantages to using a single, statewide cutoff score to assign students to remedial courses: It is quick. It is consistent. It is cost-efficient.
But is it valid? Questions were raised for me by a February Gates Foundation-funded report that Katherine L. Hughes and Judith Scott-Clayton wrote for the Community College Research Center at Columbia University. In their working paper, Hughes and Scott-Clayton describe the results of a 2009 meta-analysis (a quantitative summary of multiple studies) by the College Board, which administers the ACCUPLACER: When it comes to community college students whose test scores are high enough to exempt them from remediation, the correlation between the ACCUPLACER score and receiving a C in the relevant course ranges from .25 for the algebra exam to .10 for the reading comprehension test. (Correlations range from 0, meaning no relationship between the test and the course grade to 1, meaning the test score perfectly predicts the course grade).
Hughes and Scott-Clayton do conclude that ACCUPLACER and its ACT counterpart COMPASS are “reasonably valid predictors” at least for the students who place out of remediation. (The meta-analysis did not account for students who were assigned to remediation based upon their test scores.)
But they note that the tests are better at predicting results in math than in literacy and better at discerning which students will receive a B than which students will fail. Further, the tests do not take into account many factors that are important for college success (e.g. study skills, the presence of a strong support person). The test vendors themselves recommend using multiple measures to more accurately assign students to remediation.
An eye-opening session I attended at the recent Education Writers Association’s conference in New Orleans raised more questions for me about remediation rates based on test scores. During a panel discussion, Bruce Vandal of the non-profit, non-partisan Education Commission of the States noted that students often walk onto college campuses completely unaware that they will be required to take a test that day, much less a high stakes test that will have a profound effect on their chances of postsecondary success. Math is the most commonly flunked exam. Is it possible that a recent high school graduate has perhaps forgotten what he learned three or four years earlier in high school algebra?
For these reasons and others it is perhaps unsurprising that half of the students assigned to remediation have this reaction: They never even sign up for their remedial course.
Popularity: 61% [?]
Thanks Holly for the reminder of how remediation is determined here in CO(low-scores on the ACT, SAT or ACUPLACER). While I completely agree with you about the limitations of this indicator, a high school remediation rate when added to a school’s dropout, graduation, growth and college matriculation rates begins to get give a reasonable assessment of a high school’s quality.
There are no doubt many other important indicators of quality that should be considered but adding remediation data to the current mix would be a huge step forward. Most folks including educators are fairly clueless about how a particular school prepares its students for higher education. This seems much fairer and more robust than just relaying on 9th and 10th grade CSAP. It would also be helpful of course to see a school’s AP and IB offerings as well as their pass rates (which are currently not disclosed).
I’d be interested in knowing what data you think should be used to determine the quality of a high school.
Most concerning for me is the fact that ACT and SAT are norm-referenced assessments whose main function is to rank and sort students. This means that for any given year a 17 score in math is not consistent in showing a student’s understanding of math skills. In other words, a 17 one year might be a 18 or 19 in other years. As a high school teacher I need to know how to impact remediation rates in my school. Theoretically this could change every year.
Van-I think that all the indicators that you mentioned–including remediation rates–should be considered in the context of each high school while also acknowledging the limitations of each measure. (E.g. I think many people are aware that graduation rates are difficult to calculate. ) Other measures should also be taken into account. For instance, I know that some high schools already survey their graduates in order to solicit information on how they might improve their coursework and instruction.
I think it would also be helpful for high school teachers to have more formal opportunities to work directly with college instructors in their subject areas so as to reduce misunderstandings and improve curriculum articulation. Again, some schools already do this. Bruce Vandal also suggested that many students need more information about how the college course placement process actually works.
My main problem with remediation rates as measures of K-12 quality in Colorado is that I have recently seen many people use them as evidence of this or that without (seemingly) understanding that they are, essentially, test scores. In fact, given that the ACT is one of the three tests used to refer students to remediation, they are sometimes the same test scores that we have used for years as school accountability indicators in this state.
That said, remediation rates do provide some additional information above and beyond 11th grade ACT scores if only because they screen out students who do not try to enroll in college. Also, in cases in which students take the ACCUPLACER, they provide scores on additional tests that are not typically administered by high schools.
Holly,
Well stated. The bottom line is that ACT, SAT, COMPASS and Accuplacer have been misused by institutions and states by setting cut scores for placement in remedial education. In fact, most states don’t even use the recommended scores that indicate college readiness for these exams. Consequently these exams are insufficient for placing students in remediation. Using HS GPA combined with performance would be better, if not ideal measure.
What is exciting is that Accuplacer is coming up with new assessments that are intended to be used for placement and pinpoint academic deficiencies. This new test can be taken in HS and will be much more helpful in identifying what students need to be successful in college.
Bruce, Mark, and Holly,
I agree. First, a single “cut score” negates the whole concept of standard deviation, a concept that tells you that your “score” is valid within a _range_ (plus or minus a number of points). Second, Mark, thanks for the reminder of the difference between norm-referenced and criterion-referenced exams. If we could increase the conversations across the county among our secondary and post-secondary colleagues, perhaps we could create criterion-referenced exams (or portfolios) to demonstrate learning on a series of increasingly sophisticated levels of mastery. Could we potentially use this for advanced placement as well as placement into developmental courses? Finally, it is exciting that Accuplacer (and others, I believe) are creating diagnostic tests. I would love to see us provide diagnostic testing not only for college placement but also for rising juniors in high school.
In CA, in the CSU system, our English placement test (EPT) has been handed off to ETS , which turned it into a standardized norm referenced test. THe cutpoint for remediation is 550 on the SAT, Also there is a high correlation between SAT and EPT. No wonder that about 50% of the freshmen are tested as remedial. Might as well give them an SAT. Standardized tests are currently controlling remediation policy, with no research to validate them.
I wish we would stop relying on society’s current array of test scores as valid indicators. We are seeing the outcome of basing a child’s education on CSAP as reflected in their lack of readiness for college. We need to step back and reassess what has not been currently working in K-12 with the reliance on CSAP scores, so as to produce a student population ready for college and an educated workforce for the future. Public education can be a strong force once again if we set our priorities right and stop basing all educational outcome on the worthlessness of the CSAP and how that translates to lack of a solid, well-rounded educational curriculum. All the so-called college-readiness test scores indicate to me a lost, confused society that is mired in translating test scores while ignoring what really makes for a solid educational foundation. Let’s let teachers get back to teaching and not made to demonstrate CSAP test readiness with their students, that later is reflected as college students’ need for remediation courses.
For years, decades now, I have advocated for a process wherein a student who gets a high school diploma and yet has to take a remedial course in college have the cost/bill for that course sent to and paid by the high school awarding the diploma.
That would force articulation on measurement methods and insure rigor to the high school’s process of giving out diploma’s.
Now the high school may not prepare the student, especially in math, for college work and if having to take remedial courses the student has to pay themselves. It is as if they were deceived by the high school and have to personally pay for the problem.