Florida Board Seeks More Retention

K-12 Testing

The Florida Board of Education has asked the state legislature to mandate retention in every grade for students with low test scores. Currently, the state requires retention only for low-scoring third-graders. The proposed policy would be the most expansive in the couantry. It flies in the face of extensive research documenting the harm of grade retention, particularly for Latino and African American students, who are most at risk from such policies. More than 400,000 Florida children, disproportionately minority and low-income, would have been held back last year if the proposal had been in effect.


The debate over retention often is presented as a conflict between emotional damage and an increased risk of dropping out, on the one hand, versus improved academic achievement. This formulation is false: Retention does not improve academic achievement, even as measured by standardized tests, beyond the first year or two after retention when scores do increase slightly. Most recently, exhaustive documentation by the Consortium on Chicago School Research found that there were no academic benefits to withholding promotion, leading Chicago to reduce the use of retention (see Examiner, Spring-Summer 2004, Winter-Spring 2003).


A recent “working paper” by Jay Greene of the right-wing Manhattan Institute has been cited in support of Florida’s plan, and Greene testified before the state’s Board of Education. The report claims that in the year following the introduction of mandatory grade three test-based retention, scores rose on the Florida Comprehensive Achievement Test (FCAT). But Greene himself acknowledged, “the more important question of the policy’s effects over time is not answered in our analysis.” This acknowledgment was, however, buried in the next-to-last paragraph, far below his headline-seeking claims of retention’s benefits.


Decades of evidence shows that when students repeat a grade, their scores tend to increase for a year or two. Within a few years, those retained are back at the bottom of their class, doing no better academically than comparable students who were not retained, but twice as likely to drop out.


Greene claimed that his study is important because Florida uses “objective” criteria—test scores—to make retention decisions. However, Chicago’s policy was based entirely on test scores, as was a previous discredited program in New York City. Greene argued that retaining many children might be more effective than retaining a few - but Chicago flunked tens of thousands of children. Greene also maintained his “research” shows retention to be comparably effective to such interventions as Tennessee’s small-class STAR program, a program whose impact was evaluated over many years, not just one. Greene urged that exemptions allowing some students to be promoted despite low test scores should be sharply curtailed.


University of Tennessee Professor Dick Allington noted that half the Florida third-graders who were retained did not pass the test even after repeating a year and getting extra help. Many of those who were promoted had already failed the test twice and could not be held back an additional year. His observations in Florida, where he taught until recently, indicated that most of the vaunted extra help was provided by uncertified instructors working with large groups of children. Neither fact bodes well for the children or the program.


As is typical with his “working papers,” Greene failed to present actual data. This practice has been roundly criticized by researchers because it prevents independent analysis of the methodology that led to his conclusion. Greene has been charged by other education researchers with subordinating real research to making a policy impact based on his ideology.



• For more on Greene’s “methodology,” see Gerald Bracey’s 2004 Rotten Apples in Education Awards at www.america-tomorrow.com/bracey/EDDRA