Link to Original Story

Class Struggle
By Jay Mathews

Progress report from big study of KIPP

Mathematica Policy Research Inc. won't finish its massive study of the Knowledge Is Power Program (KIPP) until 2014, but it has issued a preliminary report. Almost all of the findings look good for KIPP, the most successful charter school network in the country, but the nation's many KIPP-watchers, including me, will not be satisfied until the research firm completes its comparison of the progress of KIPP students to similar students who applied to KIPP but were not selected in a randomized lottery.

This is the largest study ever of a single charter school network. I have promised readers I will confess when I write about KIPP that I have been studying the network for nine years and last year published a favorable book, "Work Hard. Be Nice," about KIPP co-founders Dave Levin and Mike Feinberg.

Until the randomized data comes out, many people will continue to argue that KIPP's large achievement gains are the result of the network attracting the most motivated and supportive parents. This has always struck me as nonsense. KIPP kids on average do poorly in the third and fourth grades, and much better in the fifth and sixth grades when they switch to KIPP. They had the same parents all those years. The only thing that changed was the way they were being taught. But I will shut up about that and let the social scientists do their work.

Mathematica says, in this interim report of 22 KIPP schools, "that for the vast majority of KIPP schools studied, impacts on students' state assessment scores in mathematics and reading are positive, statistically significant, and educationally substantial." Ho hum. You can find language like that in nearly every study of KIPP. But this one has some new wrinkles.

A couple of studies have suggested that KIPP may be recruiting students whose achievement level is higher than the average for children in its urban and rural neighborhoods. This preliminary study, like most others, doesn't see that. "We found no evidence that KIPP middle schools are systematically enrolling more advantaged students from their districts," Mathematica said. Since KIPP students tend to come from the poorest parts of their school districts, however, I am not sure this conclusion tells us much.

The preliminary report contradicts a notion often spread by KIPP critics on the Internet that KIPP schools have higher attrition rates than regular public schools in their areas because they are kicking out troublesome kids. Looking at these 22 schools, the researchers found that rates of attrition vary widely between KIPP schools, but average out about the same as the other public schools in their neighborhoods. By 2014 Mathematica plans to have examined 50 KIPP schools in a network that by August will have 99 schools in 20 states and the District.

Mathematica provided one statistic I had not seen before -- how KIPP's grade repetition rates compared to other schools. The study found them "consistently elevated at KIPP middle schools as compared to district public schools, particularly in fifth and sixth grades." This is expected, and a point of pride for KIPP teachers. When students arrive in fifth grade two or three years below grade level, and fail to reach what KIPP considers sixth grade level by the end of that year, they are often kept for an extra year of fifth grade. Some parents don't like this, and return their children to neighborhood schools where promoting students is the rule, rather they are ready for sixth grade or not. But the KIPP methods appears to produce more children ready for high school and college than the old methods of social promotion.

The preliminary study also revealed an unusual feature in the way Mathematica is judging KIPP school progress that will undercut another persistent criticism of KIPP data. Critics note that some studies do not take account of the fact that some of KIPP's lowest-performing students do not stay through eighth grade, making the average achievement rate of students who stay all four years of a KIPP middle school artificially high.