Date: 
Tuesday, October 25, 2011

Studies indicate charter schools performing well in reading, math

Seattle, WA - Supporters and critics alike continue to argue about whether charter schools are making a positive difference in the education of America's children. The answer, according to a new analysis of the most rigorous studies conducted to date, is a measured 'yes' in the lower grades. In higher grades, on average charter school student achievement is similar to that of students in traditional public schools.

In a new study from the University of Washington's Center on Reinventing Public Education (CRPE), economists Julian Betts and Emily Tang reviewed the 40 existing high-quality studies of charter school achievement and scientifically combined the results. The weight of the evidence from their analysis suggests that, despite considerable variation among charter schools, there is ample evidence that charter elementary schools on average outperform traditional public schools in both reading and math, and that charter middle schools outperform in math. This analysis should be viewed as more definitive than any single study because it marshals all available rigorous evidence.

The average gains from attending a charter school in one of these grades are modest but positive. For example, the largest average effect, for math achievement in middle schools, suggests that on average a charter school attendee would gain about two percentile points per year relative to their counterparts in traditional public schools. That is, a student who outscored 50 out of 100 students initially would outscore 52 out of 100 students after one year. This is not a large change, but over several years of such gains, it could be quite meaningful.

The authors found no substantial difference on average in math or reading achievement between traditional and charter high schools or in reading achievement at the middle school level.

One of the most important findings is that charter school effects vary across regions and grades quite dramatically. For instance, while overall the effect of charter schools at the middle school level is small and insignificantly different from zero, one recent study of Boston schools found very large and positive effects. The analysis also suggests that urban charters perform better than suburban or rural charters, especially at the middle and high school levels.

Students in KIPP (Knowledge Is Power Program) charter schools, according to the studies reviewed by Betts and Tang, showed significant and large improvements in both math and reading, with the estimated effects being enough to move a student initially at the 50th percentile to the 59th percentile in math and the 54th percentile in reading in a single year.

In their report, The Effect of Charter Schools on Student Achievement: A Meta-Analysis of the Literature, Betts and Tang provide a fresh overview of the currently available research. Their review focuses on only a subset of the studies, those using high-quality approaches that are more likely to yield valid results. They state that, "examining all of these results as separate parts of a whole, charter schools look to be serving students well, at least in elementary and middle schools, and probably better in math than in reading. There appears to generally be more variation in the results for math than in reading."

CRPE's associate director Robin Lake, who commissioned the report, said that "opponents and proponents are tempted to point to one or two studies that confirm their point of view. The advantage of the meta-analysis is that it looks at the entire body of rigorous research and discerns what it is saying overall. There are indeed individual studies that show charters have quality issues that need to be addressed, but on the whole, charter schools generally seem to perform well."

Do the positive results at some charter schools imply that charter students will surpass the academic performance of their traditional public school counterparts? Betts and Tang observe that, "In the long run, the variation we see in charter school achievement may shrink or grow. . . . Over time, it is possible that the number of weaker charter schools will diminish or close due to market forces, while the number of stronger charters expands."

As in their 2008 literature review (also published by CRPE), Betts and Tang note the wide disparity in the quality, and therefore the validity, of studies of charter schools. This can be a problem for journalists and others who want to know whether a given study is scientifically valid, or whether findings from one study can be generalized to charter schools overall. To address this, CRPE has published Making Sense of Charter School Studies: A Guide for Reporters.

Julian Betts and Robin Lake will discuss findings from the new meta-analysis today at the National Association of Charter School Authorizers Leadership Conference. Betts will also present the report at American Institutes for Research (AIR) on Wednesday, October 26th.