We finally know how individual schools did on the D.C. Comprehensive Assessment System tests in 2011. It turns out that some of the schools that faced questions about whether scores were manipulated in years past lost ground when tighter security was imposed this year.
We still need to know how many wrong-to-right erasures appeared on the 2011 tests compared to the 2010 tests. A case in point is J.O. Wilson Elementary School.
According to D.C. data unearthed by USA Today, Wilson had the highest portion of 2010 classrooms — 100 percent — in the city that were flagged by analysts because of higher than expected wrong-to-right erasures. (Kids rarely correct their answers. The more corrections, the more likely it is that adults changed the answer sheets.) The next three schools on the most-flagged classrooms lists were Noyes Education Campus (80 percent), C.W. Harris Elementary School (40 percent) and Leckie Elementary School (38 percent).
(Full disclosure: My wife, Linda, edited the USA Today report.)
Noyes, Harris and Leckie had some of their 2010 scores invalidated in May after investigators found signs of cheating. No violations were reported at Wilson, despite its having the largest portion of classrooms with changes of answers from wrong to right.
Why? D.C. officials don’t say. They rarely mention erasures at all. Their position is that the testing program is sound and that nearly all educators are playing by the rules.
The trouble is, we will never know what happened in schools where cheating may have occurred unless their erasure rates are examined and school staff are closely questioned in a thorough investigation.
When signs of potential cheating in Atlanta schools appeared several years ago, that school system’s leaders did what D.C. school officials are doing. They made light of the data. They released reports that absolved their schools. When the Georgia governor ordered a tough investigation run by special prosecutors, scores of educators were exposed as cheaters. Some had held test-changing parties. Some had threatened whistle-blowing teachers.
D.C. schools should release the erasure data for 2011 so we can draw comparisons to what has been revealed for 2008, 2009 and 2010. School officials should ascertain whether the lowered proficiency rates reported at several schools such as Wilson this week match a decline in wrong-to-right erasures. That would give investigators ammunition to pin down principals on what happened.
How much did the scores at Wilson decline in 2011? The portion of students scoring proficient or above in reading dropped from 66.7 to 53.4 percent. In math the drop was more severe, from 75.7 to 53.4 percent. (That’s not a typo. Wilson’s percentage of students proficient or above in reading was reported to be identical to the percentage proficient and above in math, one more oddity in a strange story.)
My colleague Nick Anderson’s story on the 2011 test results notes that of the other three most-flagged schools, Noyes also suffered a big drop in proficiency rates in 2011. Leckie had a slight drop. Harris’s results were mixed.
Why? Publish the erasure results and even I might be able to figure it out.
Maybe these tests aren’t the best way to rate our schools, but D.C. cannot depend on any assessment method until it roots out anyone who took advantage of DC-CAS. Following the Atlanta example of covering up would be the worst way to save our schools.