On Charter Effectiveness, Analysis Already Exists

To the Editor:

Jonathon Christensen and Lawrence Angel make some well-worn and exaggerated points in “What NAEP Can’t Tell Us About Charter School Effectiveness” (Dec. 7, 2005), but their arguments nonetheless have some merit. Of course, these same arguments about the limitations of National Assessment of Educational Progress data with respect to charter schools are equally true for regular public schools, private schools, and for public-private school comparisons.

How curious, then, that the authors’ methodological fastidiousness extends only to NAEP charter school results—and how fortunate that the more sophisticated analysis of charter school effectiveness that Messrs. Christensen and Angel wish for already exists.

That analysis, undertaken by the National Center for Education Statistics, applies hierarchical linear modeling, or HLM, to the 2003 NAEP charter school results and was scheduled for release more than a year ago. Even allowing for a protracted review process, the administration’s repeated delay in releasing this study is inordinate. It is also troublingly reminiscent of the inordinate delay of the official release of the 2003 NAEP charter school report, whose relatively negative findings so enraged charter school zealots.

An HLM analysis overcomes most, if not all, of the limitations of the descriptive NAEP charter school results cited by critics like Messrs. Christensen and Angel. Such critics have known about the promised HLM analysis since at least the August 2004 publication of the American Federation of Teachers’ NAEP charter school report, in which this analysis was prominently mentioned, yet their voices have not joined those calling for its release.

Regardless of one’s politics and irrespective of one’s position or nonposition on charter schools, we should all support the democratic proposition that a publicly funded report belongs in the public realm.