This means the schools may not be so different from their public counterparts. The author suggest, for further research, a randomized comparison of charter and public school students in terms of standardized state test performance to provide better insight into the question of what is 'better,' overall, for students. However, the researchers seem to be asking the wrong question: if small charter schools are specifically designed to address the needs of underserved students, why not conduct a less randomized, more specific study to see if identified at-risk or gifted students benefit more from charter schools than their counterparts in public schools? Charter schools, even if expanded, are unlikely to replace public institutions. They are a specific solution designed to address unique student needs and should be judged by specific criteria. It is not surprising that smaller schools with more individualized attention, and a need for teachers to prove their school's worth on standardized exams resulted in moderate, but statistically insignificant gains in most states. But while this may be true for the general population, for students with conditions that require specialized attention for them to function, their gains may be more impressive if they attend high-quality charter schools.

Further research is needed, but with...

If a large, randomized study is conducted, as suggested by the study authors, all students should be tested on the same test, rather than state tests that vary widely in composition. But a quantitative study that compares the effects of charter schools designed for specific student demographics might be more useful in terms of demonstrating how these schools can be used to improve the performance of the bottom or top half of American's youth. Finally, qualitative as well as quantitative data could be added to such an assessment to provide more individualized and context-specific research. Qualities that enhance learning may be very individualized, for every population group, district, even different student 'cultures.' But more definitive qualitative and quantitative research may not be fully utilizable until longitudinal studies tracking the performance of graduates are available for a wide range of different types of charter and public schools.
Reference

Greene, Jay P., Greg Foster, & Marcus Winters. (2003, July). Apples to apples:

An evaluation of charter schools serving general student populations. Center for Civic Innovation at the Manhattan Institute. Education Working Paper. Retrieved May 22, 2010 at http://www.manhattan-institute.org/pdf/ewp_01.pdf