More Musings about NAEP

Guest post by Ze'ev Wurman

Yesterday’s release of the long-term NAEP results has been already written about here, here, and here. For 9-year-olds and 13-year-olds the news is good. Significant increases since 1999, and particularly since 2004, in both reading in math. Most gratifying is the jump in reading achievement for 13- and 17-year-olds -- age groups that have been resistant to change for almost three decades.

Another piece of good news is the fact that much of the growth, in both reading and math, was concentrated in the lower achieving students, at the 50^th percentile and below. The ethnic achievement gap was either reduced in the early years or did not change because all groups kept advancing—a rising tide lifts all boats. And some of the reduction has been nothing but spectacular, as noted by Sandy Kress in his comments on Bill McKenzie's Dallas Morning News blog post. So far so good. Enhanced accountability -- whether through the standards movement of the 1990s or through NCLB in the last decade -- is finally paying off.

Much journalistic coverage has concentrated on the fact that our 17-year-olds are still essentially stuck in place, barely returning to their achievement level during the 1970s. Other coverage has pointed out that despite students taking more impressive-sounding courses, perhaps the reality is grade inflation or course-name inflation. But there is a larger question we should ask ourselves. Why should we even expect NAEP results to rise much? After all, if our education level was rather high in early 1970s when NAEP started, why should it get much better? For the 17-year-olds, there was an achievement dip in the late 1970s and early 1980s, so clearly we should be able to reclaim the 1971-73 level. But we have achieved this level in the 2008 results. Why should we expect to do notably better in the future?

One could answer that since we are doing better with our 9- and 13-year-olds, we can also expect more of 17- year-olds. Yet one could argue that the increases in the earlier grades have come not because they know more in absolute terms, but simply because now we teach material earlier, material that 30 years ago was delayed to later grades. Think of it—we push algebra to earlier grades, we expose kids earlier to a variety of cultures and facts through TV and the Internet, so they broaden their vocabulary earlier—it may be only natural that their skills look better at a fixed age. But the question still is how easily we can educate our HS graduates to a significantly higher level than in the past. Could it be that the immovability of achievement for out 17-year-olds represents some cultural constant or barrier that is hard to break, and not simply the failure of high school? It would be interesting to compare how other nations have fared with their 17-year-olds over time, but the data is hard to come by. For example, the last year of high school was sampled by the 1995 TIMSS, and it found that the best countries achieved only about 50 points above the international average, while the best countries' elementary and middle school students achieved 100-130 points over the international average. That would be consistent with the existence of some 'natural' constant that is hard to break. Makes me wonder.

[Ze'ev Wurman was a senior policy adviser in the U.S. Department of Education's Office of Planning, Evaluation, and Policy Development, 2007-2009. He also serves on the panel that reviews mathematics test-items for the California standards-based tests, and he was a member of the California State mathematics curriculum-framework committee.]