Atlanta Journal-Constitution: `Cheating' Series' Methodology is Valid
I received the following statement from the Atlanta Journal-Constitution, in response to Thursday's Educated Reporter blog about criticism of the newspaper's "Cheating Our Children" series:
"The criticisms that have emerged about our methodology are similar to those we heard from Atlanta when we first began writing about cheating in 2008 and 2009. They are similar to what the Dallas Morning News heard when they wrote about cheating in 2004. The chief complaint -- that we are identifying districts with high student mobility, not suspicious test-taking -- is not supported by the data. The data shows no correlation between high mobility and more suspicious test scores.
Statistical and erasure analyses are always simply screening tools -- they don't tell you what happened or who changed answers. As federal officials have said, it is incumbent on states and districts now to take the
"The criticisms that have emerged about our methodology are similar to those we heard from Atlanta when we first began writing about cheating in 2008 and 2009. They are similar to what the Dallas Morning News heard when they wrote about cheating in 2004. The chief complaint -- that we are identifying districts with high student mobility, not suspicious test-taking -- is not supported by the data. The data shows no correlation between high mobility and more suspicious test scores.
Statistical and erasure analyses are always simply screening tools -- they don't tell you what happened or who changed answers. As federal officials have said, it is incumbent on states and districts now to take the
