Improving the Targeting of Treatment: Evidence from College Remediation

Issue/Topic: Postsecondary Academic Affairs--Developmental/Remediation/Placement
Author(s): Scott-Clayton, Judith; Crosta, Peter; Belfield, Clive
Organization(s): Columbia University Teachers College
Publication: National Bureau of Economic Research
Published On: 10/1/2012

Besides financial aid, remedial education is perhaps the most widespread and costly single intervention aimed at improving college completion rates. The value of remediation depends not just on the average effectiveness of the treatment, but also on whether or not the individuals most likely to benefit can be identified in advance. Improving the accuracy of the placement is important, given the serious adverse consequences of incorrectly assigning a truly prepared student to remediation.

To estimate how well remedial assessments identify students who are likely or unlikely to succeed in college-level courses; to explore whether high school transcript information might be a more valuable screening device than assessments; to examine institutional trade-offs involved in establishing higher versus lower screening thresholds for remedial placement; to test whether the choice of remedial screening device has disparate impacts by race or gender.

  • Roughly one in four test-takers in math and one in three in English are severely mis-assigned. Placing prepared students in remedial courses is far more common than assigning underprepared students to credit-bearing courses.
  • Accuracy of student placement is based on the cutoff score used.
  • Using high school achievement instead of test scores to assign students results in both lower severe error rates and higher success rates (i.e., up to 10%) among those assigned to credit-bearing courses.
  • Only using high school grades disadvantages men and nonwhites. A possible policy prescription is to use both test scores and high school grades as placement measures.

Policy Implications/Recommendations:
  • Using high school transcript information for remedial assignment, either instead of or in addition to test scores, could significantly reduce the prevalence of placement errors.
  • If institutions took account of students' high school performance, they could remediate substantially fewer students without lowering success rates in college-level courses.
  • This review suggests that institutions perceive the costs of placing underprepared students in college-level courses are greater than the costs associated with placing prepared students in remedial courses.
  • Regardless of the placement tool used, 20 to 33 percent of students are likely to be severely misplaced. Policymakers should use multiple measures (e.g., test scores, high school grades, noncognitive measures) to determine placement and reduce assignment errors.

Research Design:
A predictive model that measures severe error rate (SER). Severe under-placement refers to students predicted to succeed in college-level courses but who are placed in remediation. On the other hand, severe over-placement involves the percent of underprepared students placed into college-level courses.

Unit-record data from students enrolled in a large urban and a statewide community college system.

Year data is from:


Data Collection and Analysis:
Use of an empirical model that simulates the number of students who would be placed into remediation using common cutoff rules with the two most commonly-used remedial screening tests.


Reference in this Web site to any specific commercial products, processes or services, or the use of any trade, firm or corporation name is for the information and convenience of the public, and does not constitute endorsement or recommendation by the Education Commission of the States. Please contact Kathy Christie at 303-299-3613 or kchristie@ecs.org for further information regarding the information posting standards and user policies of the Education Commission of the States.