<img height="1" width="1" style="display:none" src="http://b.scorecardresearch.com/p?c1=2&amp;c2=15743189&amp;cv=2.0&amp;cj=1&amp;&amp;c5=&amp;c15=">

'Grade deflation' hits Fairfax County

My Sept. 22, 2009, Examiner column questioned Fairfax School Superintendent Jack Dale's seemingly miraculous improvements in standardized test scores, for which he was rewarded with a lucrative contract extension. I wrote that the county's scores -- especially those among students with limited English and learning disabilities -- were being artificially inflated through the exclusion of thousands of students from the Standards of Learning exams.

In a subsequent letter to the editor, Dale characterized that column as "misleading, inaccurate and unsubstantiated." But now, the previously excluded category of students is being made to take the regular test. And in the schools where this change has happened most rapidly, SOL pass rates are down significantly.

First, a little background. In reaction to legitimate complaints that students with significant cognitive disabilities could not adequately demonstrate their grade-level knowledge on a multiple-choice exam, the Virginia Board of Education approved the Virginia Grade Level Alternative assessment for grades three through eight during the 2004-05 school year. In place of the SOLs, students chosen for VGLA could submit a portfolio of their work in SOL-tested subjects.

Unfortunately, educators soon realized that by removing low-scoring students from the larger pool of test takers, they could both increase their schools' average SOL score and their passing percentage. They also found it was easier to write up the required Individualized Education Program for academically struggling students than it was to prepare them for the SOLs. So the number of VGLA students exploded statewide.

In the VGLA program's first year, just over 2,000 students participated statewide. By the 2008-2009 school year, that number was over 47,000. By 2010, the average school district in Virginia had 20 percent of its students taking this alternative test in place of the math and reading SOLs.

A 2007 study by Virginia Commonwealth University confirmed that the "depth of knowledge" required by the VGLA was much lower than the SOLs. Educators could get test scores to jump simply by giving more children the alternative test.

In affluent Fairfax County, Dale bragged about increasing scores for students with learning disabilities and limited English by 18 to 20 points in the 2008-2009 school year. In response to a query by the Fairfax County Board of Supervisors, the school system claimed that the impact of the VGLA on overall achievement was "minimal." But between 2006 and 2009, its use in Fairfax had grown by 1,200 percent.

By early 2010, the jig was up. State Superintendent of Public Instruction Patricia Wright released a statement that "VDOE staff are concerned about the increase in the numbers of students with disabilities participating in VGLA over the last few years." The Virginia General Assembly passed a bill requiring written justification for each student's participation in the VGLA. In August 2011, Wright finally announced that the VGLA would be phased out altogether, writing that only "an extremely small number of students with disabilities" should be excluded from the SOLs.

The latest data available for the 2010-11 school year show that VGLA participation in Fairfax County's 33 low-income Title I schools has fallen by 21 percent, to just under 1,400 students. And so have reading scores.

At Lynbrook Elementary, for example, only 58 students submitted VGLA portfolios for reading last year -- down from 103 (or about 20 percent of the students tested) in the 2008-2009 year. During that same period, the SOL pass rate for the school fell from 94 percent (above state average) to 83 percent (below state average).

The same pattern can be seen in the reading scores at other Title I schools in Fairfax County. Pass rates were down in nine of the 10 schools where the number of VGLA participants fell most, and down by 8 to 12 points in six of those nine. Only seven of the 33 Title I schools improved their reading pass rates -- all seven had little or no drop in VGLA use.

We will learn more about this apparent test score deflation as the State Board of Education completely replaces VGLA with new tamperproof online testing during the 2012-13 school year. But Dale, who announced his retirement next year, will be gone before the final results are in.

Barbara F. Hollingsworth is The Examiner's local opinion editor.