As the high-profile network of schools continues to expand, KIPP leaders are taking a close look at student attrition amid arguments from critics that the loss of students at some of those public schools of choice is alarmingly high.
Attrition rates at a few KIPP schools in the San Francisco Bay Area, in particular, have recently drawn scrutiny. Fewer than half the 5th graders who entered three new middle schools in fall 2003 are still enrolled this academic year, when they would generally be finishing 8th grade, according to a KIPP analysis. At one of the schools, in Oakland, Calif., only about a quarter of the students from that 5th grade class have remained.
National attrition data on the San Francisco-based network of 52 mostly charter middle schools are unavailable. But information the network provided on a handful of other schools, as well as a review of national enrollment data by 91制片厂视频 Week, suggests that levels of student mobility vary widely across KIPP campuses.
In certain KIPP schools, in fact, attrition appears very low.
Still, some observers are raising concerns, especially given the accolades KIPP has attracted. For one, if most of the exiting students are low-performing, they say, the average test scores could be higher than they would be otherwise, and not accurately reflect the schools鈥� actual success.
鈥淭o some advocates, KIPP is the savior of public education,鈥� said Alex Molnar, who heads the at Arizona State University, in Tempe. 鈥淚f a large number [of students] don鈥檛 stay, how can we say this is a model for public education?鈥�
Yet several experts cautioned against drawing strong conclusions based on the attrition data. Student mobility, they pointed out, is high in general among low-income and minority urban families, KIPP鈥檚 prime target.
Also, they said, many of the schools are still quite new, and enrollment is likely to be unsteady early on, especially for schools of choice with the high demands KIPP has for students and families.
鈥淚 would expect to see more of that kind of attrition when schools are new,鈥� said Frederick M. Hess, the director of education policy studies at the , a Washington think tank that supports charter schools. 鈥淚 don鈥檛 see anything there that makes me doubt the value of what KIPP鈥檚 providing.鈥�
Steve Mancini, a spokesman for KIPP, said the organization is committed to keeping attrition as low as possible. 鈥淚t鈥檚 something we鈥檙e taking very seriously, trying to understand and get better at,鈥� he said.
鈥榃e Go Door to Door鈥�
KIPP has grown steadily since it was started by two teachers as a program at a Houston public school in 1994. The network now has campuses in 16 states and the District of Columbia.
Central features of the choice-based model are high academic expectations, a culture of discipline, and an extended school day that typically goes from 7:30 a.m. to 5 p.m., with Saturday classes every other week, and three weeks of class in the summer.
More than 90 percent of KIPP students are African-American or Latino, KIPP says, and more than 80 percent are eligible for free or reduced-price lunch. The schools accept students regardless of background or academic record on a first-come, first-served basis, according to KIPP.
鈥淲e go door to door in the neediest communities in the country,鈥� Mr. Mancini said.
Most students entering KIPP middle schools are two or three grade levels behind in reading and mathematics, the network reports, but quickly show strong gains, with the average student performing above grade level in all subjects by the end of 6th grade.
Some analysts argue, however, that KIPP attracts more motivated families, and that this factor and student attrition are important considerations in comparing the performance of KIPP schools with that of traditional urban schools.
鈥淭he poor kids going into KIPP are not a random cross section of all poor kids,鈥� said Bruce Fuller, an education professor at the University of California, Berkeley. And a KIPP analysis of students leaving its Bay Area campuses, he said, suggests 鈥渋t may be the weaker kids that are drifting out of KIPP schools.鈥�
Effect on Scores
An independent analysis of a rural KIPP school in Gaston, N.C., by Brent W. Maddin鈥攁 doctoral student at Harvard University鈥檚 graduate school of education鈥攍ends some credence to that hypothesis.
Looking at six entering classes at , which now serves grades 5-10, Mr. Maddin found the mean score on 4th grade tests of those cohorts was higher after the scores of students who later left were removed.
A study of attrition at a KIPP school in North Carolina found that students who left had scored lower on state 4th grade tests than those who stayed. The average baseline scores of entering students would rise if exiting students were omitted.
Change in Average Rank for 4th grade
NOTE: Math scores for the 2006 cohort were not available for the study. Results are expressed as changes in national percentile rankings.
SOURCE: Brent W. Maddin, Harvard University
That finding suggests that exiting students had, on average, lower baseline test scores than those who stayed, and thus their departure could be expected to inflate the schools鈥� average test scores. But the impact appears to taper off over time, according to the analysis.
鈥淭hat impact on test scores is much greater earlier in the school鈥檚 history,鈥� Mr. Maddin said.
His research, which has not been published or peer-reviewed, also shows the North Carolina school鈥檚 attrition has generally declined over time, with the first cohort of students having the highest level.
Jay P. Greene, an education professor at the University of Arkansas in Fayetteville, said the attrition issue is one more reason to study KIPP schools closely.
鈥淲e have a lot of very exciting test results, but we can鈥檛 entirely know whether they are attributable to KIPP instruction, or have to do with the selection of students they receive and the students they lose,鈥� said Mr. Greene, a prominent researcher on school choice. 鈥淭he ultimate remedy to these concerns is a very rigorous evaluation.鈥�
Bay Area Concerns
KIPP recently announced that, with support from the Atlantic Philanthropies, it will commission what it calls the first national, longitudinal evaluation of its schools. The independent study is expected to have a control group of comparable students who do not attend KIPP schools.
Mr. Mancini said questions surrounding student attrition will be among the issues studied. (鈥淜IPP Schools to be Studied for Long Haul,鈥� Jan. 31, 2007.)
Caroline M. Grannan, a parent activist in San Francisco, has repeatedly posted entries on the Web log that raise questions about attrition at Bay Area KIPP schools.
鈥淚 started looking more closely [at enrollment data] 鈥� and discovered this really staggering attrition at some of them,鈥� she said in an interview. 鈥淭hey鈥檙e being held up as superior to public schools while there is this huge confounding factor. That鈥檚 a problem to me.鈥�
The sizes of the 2003 5th grade classes at three San Francisco-area KIPP schools had fallen markedly by this year, for various reasons.
SOURCE: KIPP Foundation
The Bay Area schools also attracted the attention of The Washington Post, and in response to an inquiry from the newspaper, KIPP provided an in-depth analysis of attrition for three of its five schools in the area that appeared to have lost a lot of students. The network also made the analysis available to 91制片厂视频 Week.
The KIPP analysis, which relied on exit interviews with families conducted by school staff, suggests that about half the departing students moved out of the area, while the other half chose to leave for reasons mainly tied to the KIPP approach. Only two students at one school, KIPP says, were explicitly told to leave.
For the students who did not move, the most common reason cited for leaving was to avoid repeating a grade. KIPP schools are aggressive about promoting students only if they demonstrate the knowledge and skills needed to advance. Other common reasons cited were the extended school day and the strict discipline code.
being conducted by SRI International, a Menlo Park, Calif.-based research outfit, also offers limited analysis on the reasons students leave.
鈥淪ome leave to avoid retention,鈥� said the first-year report, issued in March 2006. 鈥淥thers leave because they move, have needs that the school cannot meet (severe learning disabilities, emotional needs, or behavioral problems), or because there is a lack of commitment on the part of students and/or parents.鈥�
Interviews with KIPP staff members, the SRI study said, suggest that in general the schools do not push students out.
Rates Drop Over Time?
Data the KIPP Foundation provided 91制片厂视频 Week from a sampling of five schools show some preliminary evidence that attrition rates drop over time.
A campus in Newark, N.J., for example, has seen a steady decline in annual attrition, from about 26 percent during its first year to 8 percent four years later. At KIPP Bridge College Preparatory in Oakland, which showed a high level of attrition for its fall 2003 cohort of students, early data indicate the annual exodus is going down.
鈥淎s schools become more established, there鈥檚 some evidence that the schools are doing things that are leading to more students鈥� staying with KIPP on an annual basis,鈥� Mr. Mancini said. 鈥淭hat data is very promising.鈥�
Attrition rates seem to vary widely across the network, according to the data KIPP provided and an 91制片厂视频 Week analysis of enrollment at 23 KIPP schools nationwide.
For instance, a KIPP pre-K-elementary school in Houston, housed on the same campus as one of the first KIPP middle schools, has seen annual attrition of less than 4 percent in its first three years, the network鈥檚 data show. KIPP Ways Academy, in Atlanta, saw annual attrition fluctuate over its first three years, from 21 percent up to 30 percent, then back down to 21 percent.
Data from the Fulton County, Ga., district show that South Fulton Academy, in East Point, had 75 5th graders in 2003-04 and roughly half that number of 8th graders鈥�39鈥攖his academic year. Those figures do not account for students who repeated a grade or students who entered the school later, though KIPP schools typically add few students in the upper grades.
In sharp contrast, two Houston middle schools鈥擪IPP Academy Middle School and KIPP 3D Academy鈥攈ad 8th grade enrollments last school year that were 95 percent or more of the 5th grade class three years earlier, state data show.
鈥楽erve Every Student鈥�
KIPP principals say they are communicating more across the network about strategies to keep more students, and to learn from those who succeed.
鈥淲e鈥檙e going to do everything we can to serve every student who comes through our doors,鈥� said Molly H. Wood, the principal at Bayview Academy in San Francisco. 鈥淲e are getting better at explaining why we do what we do, and why the high standards are worth it.鈥�
For instance, she said, it is critical to assure families that repeating a grade may be the key to success. 鈥淔or 5th grade this year, we鈥檙e planning to hold eight kids back, and at least seven of those families are completely on board,鈥� she said.
David R. Ling, the principal at the Oakland KIPP school, says there鈥檚 not much he can do about high family mobility in the area his school serves. Data for the Oakland Unified School District show that, on average, about half the students who enroll in middle school in 6th grade districtwide either change schools or drop out before the end of 8th grade.
But he said he鈥檚 working to convince families that the KIPP demands will pay off.
鈥淲e鈥檝e always told parents,鈥� he said, 鈥淜IPP is going to be harder than your old school.鈥�
