Ildiko Laczko-Kerr and David C. Berliner, Education Policy Analysis Archives
September 6, 2002
The title and subtitle will give you a sense of the objectivity of the so-called on-line "peer-reviewed scholarly journal" named Education Policy Analysis Archives, edited by Gene V. Glass. Like most such journals in education, it obviously selects peers who agree with its editors' policy biases and those of the authors whom they like and wish to publish. In this case, the authors are Ildiko Laczko-Kerr, who currently works at the Arizona Department of Education, having recently completed a doctorate at-where else?-Arizona State University, where her co-author, David C. Berliner, is a bigfoot in the College of Education and where this "journal" is "published." (The article is derived from Dr. Laczko-Kerr's dissertation, entitled "Teacher Certification Does Matter.") The early pages contain a thoughtful review of some previous research on the interaction of subject-matter knowledge and pedagogical prowess in influencing the effectiveness of public-school teachers. Not too surprisingly, they conclude that teachers tend to be more effective when they know something about what they're teaching, and that knowledge becomes more consequential as grade levels rise. The new "research," however, leaves much to be desired. The authors picked five Arizona school districts with teacher shortages, mostly low income communities, then selected "matched pairs" of certified and "under-certified" teachers who began work in their public schools in 1998 or 1999. (Of the "under-certified" group, two-thirds held "emergency" certificates which, it is commonly realized, have nothing to do with programs like "Teach for America" but simply getting a "warm body" in front of the class. Of the certified teachers in this study, incidentally, half were products of Arizona State.) The authors compare the test scores of the students of the certified teachers with those of the "under-certified" teachers and find that the students of the certified teachers have higher scores. Since there are no controls in the study for the prior test scores of the students, the researchers are on extremely shaky ground in drawing conclusions about the "value added" by the two groups of teachers. It is entirely possible that "under-certified" teachers were assigned to classrooms with students whose prior level of achievement was below that of other students. The authors attempted to control for differences across classrooms by matching teachers in similar schools or districts, but they are unable to match classrooms on the most important variable: prior student achievement. The authors themselves admit that some of the matches they made across districts were flawed. If the authors had used longitudinal student achievement data for their study, they would have been able to control for prior student achievement and to draw conclusions about the effectiveness of the different groups of teachers. If you want to evaluate this study for yourself, you can find it on-line at http://epaa.asu.edu/epaa/v10n37/.
