Another Way to Game the System

Abstract

BACK in 1977, when I had just taken a job directing Virginia's testing programs, state was in an uproar over actions of one school district. At time, Virginia administered an test along with its norm-referenced achievement test in grades 4, 8, and 11. State board of education policy held that a district's achievement should be at least as high as its ability. It was a pretty naive conception, but it reflects a widely held but erroneous notion that ability tests measure potential and achievement tests measure extent to which schools have succeeded in bringing their charges to that potential. The previous year, one of 141 districts in state had been singled out for failing to produce achievement that ability test said it should. In response, superintendent had retained in grade 67% of his third-graders. Predictably, fourth-grade achievement scores soared. According to a recent analysis by Walt Haney of Boston College, is now perpetuating a phony view of progress using same technique. In 2005 showed impressive gains on grade-4 mathematics test of National Assessment of Educational Progress (NAEP). While nation showed blacks up four points and whites up three, showed blacks gaining nine points and whites four. fourth-graders moved slightly ahead of fourth-graders nationwide. Gov. Jeb Bush teamed up with New York City Mayor Michael Bloomberg to tout their progress in a 13 August 2006 Washington Post op-ed. Florida and New York City are leaders when it comes to accountability in education. We have set high expectations for all students, and in key grades we have eliminated promotion, harmful practice of pushing unprepared students ahead. A more cliche-ridden essay is hard to imagine, but what we have here, at least in Florida, is another case of skim milk masquerading as cream. Haney calculated what he calls for and nation. These ratios involve dividing number of students enrolled in one grade in one year by number enrolled in previous grade in previous year. What you get is an index of proportion of students promoted or retained. For example, if you have 100 second-graders in 2004-05 and 125 third-graders in 2005-06, transition ratio for third grade will be 1.25. Nationally, ratio is about 1.07 for first grade and very nearly 1.00 for other grades through grade 8. Many districts take a hard look at promoting students from first grade, which explains higher ratio at that level. Those who are promoted from first grade then progress through grades with very little retention until ninth grade. At ninth grade, ratio is around 1.13, reflecting the ninth-grade bulge. Over five-year period of Haney's analysis, there is little change in any of ratios at national level. Florida's statistics look different. First, at lower grades, almost all ratios exceed 1.00. Haney states that this reflects in part growth in number of school-age kids. Between 1999-2000 and 2003-04, added about 300,000 students K-12. The transition ratios Haney calculated for for four years in 12 grades appear in Table 1. Perhaps most dramatic data in this table are those for ninth grade, which show that is retaining something like 30% of its ninth-graders, about double national average. Similarly, very low numbers at 10th grade indicate an enormous attrition between ninth and 10th grades. Florida's abolition of social promotion has certainly jacked up its early dropout rate. Second, and bearing more on Bush/Bloomberg contention, transition ratios are consistent for four years shown except at third grade for 2003-04. The year before showed its great progress on NAEP math test (2004-05), state retained many more third-graders. Elsewhere in his article, Haney indicates that only 4% to 6% of white third-graders were flunked, while 15% to 20% of blacks and Hispanics were held back. …