Charter study uses wrong metric for high performing


In the last few days, the Center for Research on Education Outcome at Stanford University (CREDO), published a report detailing education outcomes for students in charter schools across 27 states (including MN), covering 95% of all students in charter education. The broad results were mixed, aggregating huge numbers of school test results into single numbers pitting all charter schools in a state against a virtual traditional counterpart. Even when these results arebroken down by demographic, these kind of broad, test-based results do not truly draw conclusions on the quality of education being offered by any type of school.

The biggest issue at play whenever education quality is assessed is the problem of trying to figure out what to measure. In this study, CREDO has only looked at test results, which by itself is rather problematic. This may provide the closest comparison statistically, but it does very little to illuminate what is actually going on in the school.

There is also an issue in the way CREDO attributes charter performance increases from the 2009 study to 2013 study. Its press release says the gain is due to the closing of underperforming and the opening of more high-performing charters. It further states that the charter movement wants to continue this trend. On the surface, this seems logical. But here’s what’s troubling. Even if a charter school is doing well with particularly at-risk students, the raw test scores may not reflect this success. This increases pressure on charter schools to only cater to students who already demonstrate an ability to succeed, leaving the rest for traditional schools. This may make the charter schools look better, but does not truly address any problems currently at hand.

The problems facing education right now are numerous and complex, and there is not going to be any magic formula to solving that problem. By offering alternative choices, charter schools should be trying to reach out to those for whom the traditional model is not working. However, these students are also less likely to perform on standardized testing, and reports like this one create an environment that discourages focusing on these students. Broad generalizations on education are dangerous, and this CREDO study shows that.