Audrey Amrein and David Berliner, Education Policy Analysis Archives
March 28, 2002
Audrey Amrein and David Berliner of Arizona State University are the authors of this 70-page cyber-article that seeks to determine whether high-stakes testing has actually boosted student achievement in eighteen states that, say the authors, have attached "severe consequences" to their testing programs. They don't seek evidence from the states' own tests but, rather, from other national data sources. They conclude that high-stakes testing doesn't boost student achievement, that gains shown on state tests are the result of various manipulations (e.g. curricular narrowing, excluding students), and that undesirable consequences are rampant. The essay also serves as another opportunity for Berliner to restate his familiar view that the nation is not and never was "at risk." The piece, overall, is more hatchet job than careful social science. The information they use about many of the eighteen states in their sample is not entirely accurate. For example, they ascribe to a number of states "consequences" that haven't yet taken effect or have affected only a tiny number of students or schools. Their principal sources of external data are college-admissions tests (SAT, ACT, AP) that are not taken by all students and that are less apt to be affected by state-level accountability policies aimed at low-performing students and schools (those being students less apt to even apply to college); and NAEP results, which are not even available at the state level in 12th grade. What sort of social scientist would use NAEP 4th and 8th grade scores to (in the authors' words) "test the effects on learning from using high-stakes tests in states that have implemented high-stakes high school graduation exams"? In sum, I see no basis for accepting the authors' starting premise that "the ACT, SAT, NAEP and AP tests are reasonable measures of the domains that a state's high-stakes testing program is intended to affect." To my eye, they fail to make their case. If, nonetheless, you'd like to see for yourself, surf to http://epaa.asu.edu/epaa/v10n18/. If you'd like to see more systematic rebuttals of the claim that tests don't boost student achievement, by Cornell's John Bishop and colleagues, you can find two of them (in PDF format) at http://www.ilr.cornell.edu/depts/cahrs/PDFs/WorkingPapers/WP00-09.pdf and http://www.ilr.cornell.edu/depts/cahrs/PDFs/WorkingPapers/WP98-27.pdf. You may also want to check out the 2001 edition of Brookings Papers on Education Policy, edited by Diane Ravitch, about which you can learn more at http://www.brookings.edu/dybdocroot/press/books/bpep2001.htm.
