Why did Lucille Davy and the DOE suddenly raise the score required for “proficiency” on N.J.’s standardized assessments for middle schoolers after the tests had already been graded? Same reason they did it last year for 3d and 4th graders: New Jersey’s standardized assessments wildly overstate our kids’ academic skills when compared with the gold standard: the National Assessment of Educational Progress.

This set of tests, commonly known as the NAEP, is given to samplings of kids across the country at 4th, 8th, and 12th grades and provides a common metric so that one state can really see how its kids do compared to other states. An analysis of N.J.’s results (here’s all the data) shows that, indeed, our state tests – the ASK’s, the GEPA’s, and the HSPA’s – artificially inflate our performance. In addition, the NAEP results highlight the achievement gap between higher income kids and poor kids.

In 2007, our state-designed NJ ASK test given to 4th graders resulted in 81% of them deemed proficient in reading. The NAEP test given the same year showed that 43% of our 4th graders read proficiently. In other words, our state test overestimates by a factor of 2 the percentage of our kids who are able to successfully perform on the national assessment. When broken down into subgroups, the NJ ASK test results labeled 61% of our African-American kids proficient in reading; the NAEP showed 18%. Compared to the other 49 states, though, our results aren’t too shabby: we came in 4th in for African-American kids and 5th for Latino kids. Socio-economic class differences were stark. The reading performance of our 4th grade higher income kids was 3d in the country and our lower income kids came in 16th.

Let’s jump to 8th grade math, where N.J. uses the GEPA. 69% of our kids were deemed proficient in the state-designed GEPA, though only 40% of 8th graders were proficient in math when assessed by the NAEP. While our higher income kids came in 4th in the country on Grade 8 mathematics, our lower income students came in 28th. We just beat Missouri and Arkansas. The NAEP also looks at access to qualified teachers, certainly a factor in student achievement. How’d we do? Across the nation, we came in 41st in access to teachers who are trained in the field they instruct.

As far as cost per pupil, we’re 1st in the country.

It’s not all bad news. From 2003 to 2007, the performance of African American kids in 4th grade reading went up by 12 points, though Latino kids only went up by 2 points. In grade 8 math, African-Americans went up 11 points over 4 years, Latinos went up 9 points, and Whites went up 6 points. Compared to the rest of the country, we were 15th in math performance for African Americans and 9th in math performance for Latinos.

What really jumps out is not the discrepancy in performance among ethnic groups, but the discrepancy in performance between higher income and lower income kids. If anyone needs more proof that the N.J. public education system is inconsistent and segregated, it’s all right here.

At a NJEA conference last year, Commissioner Lucille Davy was asked why the State DOE raised proficiency levels on NJ ASK scores after they had already been graded. She replied, “In my heart, I thought it was right to raise proficiency standards. We must prepare our kids for success in the 21st century.” Well, it’s not just her heart telling her it was right; it’s the NAEP telling us that our State assessments are wrong.