KIPP attrition vs SFUSD attrition
I'm only doing this by popular demand (see comments on previous post about KIPP attrition). I'll bet KIPP wishes you wouldn't egg me on.
OK, San Francisco's two KIPP schools are distinctive in that they do not fill up for grade 5, which is the starting grade for all KIPP schools (not counting a new high school or two). Jay Mathews, the Washington Post/Newsweek education writer who covers KIPP frequently and admiringly, asked me about that, because apparently it's unusual for KIPP. Evidently students in SFUSD K-5 schools, and their families, aren't eager to switch until it's time to move on to middle school — kids don't want to miss their 5th-grade graduation. But then the San Francisco KIPP schools get a surge of 6th-grade applicants, and KIPP says both have 6th-grade waiting lists (and KIPP confirms that the 5th grades do not have waiting lists).
So because of that I'm looking at the enrollment from 6th to 8th grade for the most recent 8th-grade class, which finished grade 8 in June '08. Note that these figures (including SFUSD's) are for FALL of 8th grade and do not indicate how many finished 8th grade. The attrition in the two KIPP schools was significantly higher than attrition in SFUSD.*
*Note that it only makes sense that attrition in the two KIPP schools would be much higher than attrition in the district overall, because students who leave the KIPP schools are likely to move to a non-KIPP SFUSD school, while mobile students within SFUSD still show up in the district statistics.
OK, San Francisco's two KIPP schools are distinctive in that they do not fill up for grade 5, which is the starting grade for all KIPP schools (not counting a new high school or two). Jay Mathews, the Washington Post/Newsweek education writer who covers KIPP frequently and admiringly, asked me about that, because apparently it's unusual for KIPP. Evidently students in SFUSD K-5 schools, and their families, aren't eager to switch until it's time to move on to middle school — kids don't want to miss their 5th-grade graduation. But then the San Francisco KIPP schools get a surge of 6th-grade applicants, and KIPP says both have 6th-grade waiting lists (and KIPP confirms that the 5th grades do not have waiting lists).
So because of that I'm looking at the enrollment from 6th to 8th grade for the most recent 8th-grade class, which finished grade 8 in June '08. Note that these figures (including SFUSD's) are for FALL of 8th grade and do not indicate how many finished 8th grade. The attrition in the two KIPP schools was significantly higher than attrition in SFUSD.*
KIPP S.F. Bay Academy:
Total enrollment for that one class:
05-06 (6th grade) 75, 06-07 (7th grade) 55, 07-08 (8th grade) 44. That's a loss of 41.4% of the class.
This KIPP has enough of a Latino population that I crunched those numbers too, as well as African-American. The subgroup with the highest attrition was African-American girls, which is unusual; in other KIPP schools the subgroup with the highest attrition has been African-American or Latino boys. Please let me know if I've miscalculated any percentages.
African-American boys in that one class:
05-06 (6th grade) 17, 06-07 (7th grade) 15, 07-08 (8th grade) 9. That's a loss of 40% of the African-American boys in the class.
African-American girls in that one class:
05-06 (6th grade) 21, 06-07 (7th grade) 14, 07-08 (8th grade) 10. That's a loss of 52.4% of the African-American girls in the class.
Latino boys in that one class:
05-06 (6th grade) 9, 06-07 (7th grade) 7, 07-08 (8th grade) 5. That's a loss of 45% of the African-American boys in the class.
The number of Latino girls just went from 11 to 10, so I didn't bother to crunch that subgroup.
KIPP Bayview Academy:
Total enrollment for the class that finished 8th grade in June 2008:
05-06 (6th grade) 88, 06-07 (7th grade) 58, 07-08 (8th grade) 47. That's a loss of 46.6% of the class.
African-American boys in that one class:
05-06 (6th grade) 27, 06-07 (7th grade) 27, 07-08 (8th grade) 10. That's a loss of 63% of the African-American boys in the class.
African-American girls in that one class:
05-06 (6th grade) 32, 06-07 (7th grade) 34, 07-08 (8th grade) 20. That's a loss of 37.5% of the African-American girls in the class (after a small increase from 6th to 7th).
San Francisco Unified School District:
Total enrollment for the class that finished 8th grade in June 2008:
05-06 (6th grade) 4,106, 06-07 (7th grade) 4,025, 07-08 (8th grade) 4,016. That's a loss of <3%.
There was striking attrition for African-American boys and African-American girls in that grade, districtwide -- but far below that of both KIPP schools.
African-American girls in that one class:
05-06 (6th grade) 320, 06-07 (7th grade) 230, 07-08 (8th grade) 266. That's an overall loss of 16.9% of the African-American girls in the class.
African-American boys in that one class:
05-06 (6th grade) 318, 06-07 (7th grade) 233, 07-08 (8th grade) 287. That's an overall loss of 9.75% of the African-American boys in the class.
Attrition of Latino girls was >4%. The number of Latino boys in that one class rose from 456 in grade 6 (05-06) to 465 in Grade 8 (07-08).
*Note that it only makes sense that attrition in the two KIPP schools would be much higher than attrition in the district overall, because students who leave the KIPP schools are likely to move to a non-KIPP SFUSD school, while mobile students within SFUSD still show up in the district statistics.
Labels: Charters, SFUSD Politics
6 Comments:
Interesting information.
I think your final point is very important when discussing the reality of the KIPP attrition rate: "It only makes sense that attrition in the two KIPP schools would be much higher than attrition in the district overall, because students who leave the KIPP schools are likely to move to a non-KIPP SFUSD school, while mobile students within SFUSD still show up in the district statistics."
I'm wondering how you accounted for the fact that regular SFUSD schools enroll students at all grade levels? KIPP does not accept new students in seventh and eighth grades, so any students that move on are not replaced.
I'm sure KIPP wishes that all students would stay with and benefit from the program, but I don't see any data showing that their attrition rate is particularly high. Am I missing something?
I didn't account for that difference. I just compared the two sets of attrition figures after being urged to by posters.
On its face, attrition at these KIPP schools is particularly high -- you DO see the data, right on my post. I looked at attrition figures for all 9 California KIPP schools (they may have a new one; I did this a year or so ago). 6 of the 9 showed the same pattern: strikingly high attrition -- and far higher for the subgroup that is statistically likely, on average, to be the most academically challenged.
The attrition at district non-KIPP schools could be just as high, but the difference is that (as you observed) KIPP does not replace the students who leave in higher grades. (Actually, KIPP will apparently enroll some students in grade 7 -- I inquired about enrolling my own 7th-grader, and they said they would test her and then enter her in the lottery for the grade she tested into. But they say they enroll few 7th-graders and no 8th-graders.)
The point is that it's (on average) the more challenged, struggling students who are likely to leave, for whichever reason*. High-mobility students are statistically likely to be more academically challenged. If one of those students leaves a traditional public school, the way the system works is that another student -- equally high-mobility -- is expected to replace him or her. If one of those students leaves the KIPP school, he or she is not replaced, and the school has one fewer academically challenged student (based on the likelihood, on average, that the high-mobility student will be a lower achiever).
So that is obviously likely to have a positive impact on KIPP's achievement. When such a very high percentage of students leaves and is not replaced -- and an even higher percentage of the subgroup that is statistically likely to be the most academically challenged -- and those students are not replaced, that's likely to have an extremely significant positive impact on KIPP's achievement.
KIPP itself has responded to my findings by expressing concern and implementing studies of its own attrition, so while some less-informed outsiders who admire KIPP might try to dismiss the findings as insignificant, KIPP itself clearly does not agree.
*It would seem likely that some students would leave KIPP because they couldn't handle the rigorous academics and/or discipline; others might be likely to leave because KIPP wanted to hold the student back to repeat a grade. A cynic would suspect that some challenging students might be pushed out -- "counseled out" is the usual term -- though KIPP says this is not its practice. When individual principals are under pressure to show results, it would seem somewhat likely that this might happen. Obviously, any or all of these things would result in KIPP's shedding lower-achieving students and keeping the higher achievers, while not replacing the departed lower achievers.
(Sorry, realized how rude it sounded to imply that a legitimate commenter was uninformed; didn't mean to be so heavy-handed.)
I'll go blind if I crunch all the school numbers, but I looked at SFUSD's current lowest-performing middle school, Horace Mann. For the class that finished 8th grade in June '08:
Total class: 6th grade (05-06) 181, 7th grade (06-07) 172, 8th grade (07-08) 176.
Latino boys: 6th grade (05-06) 72, 7th grade (06-07) 68, 8th grade (07-08) 73.
Latino girls: 6th grade (05-06) 71, 7th grade (06-07) 64, 8th grade (07-08) 71.
So now you get the picture, right? A high number of kids probably leaves Horace Mann -- perhaps as high as the number who leave KIPP -- but at Horace Mann they are replaced with similarly high- mobility kids. So the mobility doesn't have the same likely effect, which is to get rid of the likely lower achievers.
It should be possible to account for KIPP's policy of not replacing kids who leave. But it would not be as simple as computing an average attrition rate for each grade and compounding that accross the three years. Since KIPP's policy means that high-mobility kids will vanish and not be replace by other high-mobility students, one should see lower attrition rates in the upper grades. Over time you would expect to see a less-mobile, more stable set of students with lower attrition. We don't see that do we?
If it turns out that one of KIPP's 'advantages' is that is concentrates the more stable students and families in one school, it might be interesting to try this in the public middle schools. Designate one or two 'academies' among the middle schools that make it clear that vacancies will not be filled -- that the school wants to promote a stable learning environment where the students grow up together and learn to work together.
I didn't see anyone asking you to compare the attrition of a couple KIPP schools to the attrition of the district. I did see people asking you to compare KIPP schools overall to their districts. You're good at pointing at two schools and extrapolating a false conclusion from them, but you haven't proven to be very good at using enough data to draw valid conclusions.
The sad reality is that "dumping" and "shuffling" does happen prior to STAR testing in San Francisco schools, and KIPP is no exception. As KIPP cannot receive transfers, but can force students out, non charter SFUSD schools have the responsibility to educate students that either couldn't make it or were rejected by KIPP. These students test scores are not reflected on KIPP's numbers but on the school where the students are transferred to. A normal SFUSD school cannot reject any student. In a hypothetical (and cynical) situation, when a student is transferred a week before testing from KIPP to a neighboring school, the neighboring school picks up that student's test scores in their stats, while KIPP avoids them, helping their numbers stay high, despite the fact that the student was educated by the staff at KIPP, not the new school. Perhaps this accounts for some percentage of the attrition.
Post a Comment
<< Home