This today from the Core Knowledge Blog in reference the discrepancy between New York State testing in math, which showed that 80% of 8th graders reached proficiency, and NAEP results that showed that 34% of 8th graders reached proficiency. Much rending of garments, shouts of “foul” from defenders and detractors of Mayor Bloomberg and Commissioner Joel Klein.
In New York, 80 percent of 8th graders met the state’s standards in math this year, up from 59 percent two years ago. But the National Assessment of Educational Progress (NAEP) results released yesterday paint a different picture. Only 34 percent of the state’s 8th graders are considered proficient, a modest increase from 2007 levels. NAEP scores for the Empire State’s 4th graders actually declined, while the percent passing the state’s own test went up. This renewed charges that New York is making itself look good by lowering standards. Diane Ravitch puts it plainly: ”The fabulous ‘gains’ reported last spring, we now know, were based on dumbed-down tests and dubious scoring of the tests in Albany,” she writes in today’s New York Post.
We’re in a similar pickle in N.J., with gaping disparities between our state tests for 4th and 8th graders, NJASK4 and NJASK8, and the national test, NAEP. For example, our state tests say that 85% of our 4th graders are proficient in math. The NAEP says that 49% of our 4th graders are proficient in math.
Here’s the ethnic breakdown:
In 4th grade, according to the NJASK4, 92% of our white students are proficient in math, 68% of our black students are proficient in math, and 76% of our Hispanic students are proficient in math. But according to NAEP data, 63% of our white students are proficient in math, 19% of our black students are proficient in math, and 25% of our Hispanic students are proficient in math.
For 8th graders, (based on 2007 data because the State correctly raised the cut score for proficiency in 2008, skewing the numbers), NJASK8 results show that 81% of our white students are proficient in math, 38% of our black students are proficient in math, and 51% of our Hispanic students are proficient in math. According to NAEP data, however, 54% of white students are proficient in math, 17% of black students are proficient in math, and 22% of Hispanic students are proficient in math.
We’ll leave it to you statistics wonks out there to figure out standards of deviation, but the discrepancies are striking, certainly on a par with those in N.Y. Yet, in the midst of a gubernatorial race which seems to spare no rods, the response has been muted, almost blasé. Governor Corzine and the DOE did put out a press release extolling N.J.’s fabulous gains – “[The state test scores] are a clear sign,” said Commissioner Davy, that our approach to improving teaching and learning is working and that the investment in our schools is paying off” – but if there’s been any gnashing of teeth ala Diane Ravitch in the Post, then we haven’t seen it.
To the DOE’s credit, they are raising cut scores this year for 3d, 4th, and 5th graders. But discrepancies of this magnitude are not mitigated by tweaking definitions of proficiency. It’s a sound case for national standards and national testing. The NAEP is viewed as a low-stakes sideshow. Maybe it should be the main event.
Correction: A sharp-eyed reader notes that the DOE raised cut scores this year on the NJASK 3 and NJASK 4. The 5th grade cut-off for proficiency was raised last year.
Here's a link to a really nice NCES report which shows the relationship between state test cut scores and NAEP scores. In general, NJ cut scores for proficiency, while lower than NAEP cut scores, are in the upper middle portion of the pack for 8th grade, but lower for 4th grade.
Makes you wonder about the value of state tests.
Thank you for posting this. Our state performance on the NAEP is still failure. This is a national epidemic. We must take this seriously for our children's futures are at stake.