CCRC Publishes Studies Critical of Placement Testing

On February 28th, the Community College Research Center (CCRC) at Columbia University released two studies focused on the use of standardized tests to determine placement in developmental education at community colleges which were widely publicized in the New York Times and other media outlets.  http://www.nytimes.com/2012/02/29/education/colleges-misassign-many-to-remedial-classes-studies-find.html?_r=1&emc=eta1

Using data from a large urban community college system, the first study found misplacement rates of 24 to 33 percent using COMPASS.  The second study, using data from a statewide community college system, found that three out of every ten students may be incorrectly assigned to either developmental or college-level courses using COMPASS and ACCUPLACER.  Both studies found that in general placement tests are more accurate for math than for English.  Both studies also indicated that placing students based on high school GPA would yield a lower percentage of severe placement errors than the tests.  The researchers stopped short of recommending that colleges discontinue using the tests but did suggest alternatives that could yield better results.

In fall 2009, the most recent year for which data is available, an average of 62% of incoming students placed into one or more developmental education courses at Michigan’s community colleges (http://www.michigan.gov/midashboard/0,4624,7-256-58084---,00.html). This rate includes both recent high school graduates and adults entering college for the first time.  Despite the stated intent of developmental education to prepare students with weak academic skills to succeed in college, recent research has demonstrated that students assigned to these courses frequently fail to persist and have lower levels of overall educational attainment than students who begin at college level.  These two new CCRC studies now strongly suggest that the most commonly used placement tests may not be accurate for significant numbers of students.  Based on this research, some policy and practice recommendations for consideration by Michigan community colleges include:

1)      Collaborate with K-12 partners to share information about college readiness standards and how both high schools and colleges work with students to help them meet those expectations.  Implementation of the Common Core Standards and more rigorous proficiency scores on the MEAP and MME demonstrate commitment on the part of the Michigan Department of Education to graduate more students who are college and career ready.  The MCCA and the principals’’ association (MASSP) have recently agreed on common college readiness benchmarks based on a range of ACT scores in Mathematics, Reading and Writing that will allow students to bypass placement into developmental education. Additionally, partnerships to increase dual enrollment, cross-sector faculty and administrator dialogues and more effective use of information from assessment tools including EXPLORE, PLAN and ACT can help reduce the numbers of recent graduates who are placed in developmental education. 

2)      Examine the conditions under which placement tests are administered on campus.  Colleges advise students to complete placement tests prior to meeting with an advisor for their initial registration.  Since community college students typically have limited time and access to transportation, many students try to complete testing, advising and registration within one campus visit, often immediately before the semester begins.  Providing advising and access to test preparation resources well in advance of having students take the tests as some colleges in Michigan and other states are doing through their orientation programs may decrease the number of students referred to developmental education who could have succeeded in college-level coursework.

3)      Consider additional measures beyond standardized tests.  Scarce resources of money, time and personnel have prompted colleges to rely on these tests, which are designed to be quick and inexpensive to administer (each individual test costs less than $2.00 and takes about ½ hour to complete).  However, including or substituting measures such as high school GPA, high school attended, years since high school graduation and assessment of other non-cognitive factors such as the GRIT assessment being piloted by Grand Rapids Community College may yield better placement results resulting in higher levels of credit and credential attainment over time.

4)      Develop alternatives to a “one size fits all” approach to placement in and delivery of developmental education.  Colleges typically rely on a single score (perhaps two in math) to determine who is “college ready.”  This generic distinction ignores the fact that different types and levels of mathematics may be needed in Culinary Arts versus STEM disciplines or that more sophisticated rhetorical skills may be required to complete a Social Science versus a Welding degree.  By examining learning outcomes from each program and the skills needed to meet them, states such as Virginia and North Carolina are implementing  modularized, discipline-specific pathways leading from developmental through college-level  mathematics as well as diagnostic assessments that assign students only to the modules they will need to complete their programs.  Contextualizing developmental reading, writing or mathematics content within program-specific instruction is another approach that has shown promise in Washington and other states.

Implementing any or all of these recommendations will require investments of money, time and personnel and most will increase the complexity of admissions and enrollment for students.  In addition to concern over the resources required, colleges who have worked diligently to simplify processes at their front doors in order to improve access may justifiably question whether their students will perceive requirements to attend a mandatory orientation, present their high school transcript or participate in a longer, more intensive advising procedure as insurmountable obstacles to enrolling in college.  However, research conducted by the Center for Community College Student Engagement at UT Austin strongly suggests that students would prefer that colleges not only offer but require more guidance for entering students as they select a program of study and register for courses.  In addition, reallocating resources to increase assistance to entering students has the potential to generate a return on that investment through increased persistence into higher level courses and increased credential completion.