Thursday, October 29, 2009

Attendance area preferance

PPSSF asked me to post the following:

There has been a lot of discussion and confusion regarding the attendance area preference in the Student Assignment System and where you should rank that school on the application I just received confirmation from the EPC that the statement below is how the Student Assignment System treats attendance area schools.

In the technical description of the SAS it states that for attendance area schools, attendance area kids within the applicant pool will be selected as long as their demographic profile increases diversity. Further it states, that once attendance area kids no longer increase diversity and all kids within the applicant pool are considered, if there are multiple kids within a selected demographic profile than the kid who lives within the attendance area will be selected for assignment. It does NOT state that this attendance area preference is only considered if the attendance area school is the applicants first choice. The technical description of the SAS can be found at http://portal.sfusd.edu/data/epc/DI_Handout_Combo.pdf.

Bottom line is that attendance area preference is given regardless of the school rank on the application and it does not require the applicant to place their attendance area school first on their application.

Vicki Symonds
Parents for Public Schools-SF
vicki@ppssf.org

78 comments:

  1. Vicki, thank you for tracking this and posting a clear statement about it.

    ReplyDelete
  2. I was the parent who asked Rachel Norton about this on a different thread. Thanks to Rachel and Vicki for the follow-up!

    ReplyDelete
  3. What about families who reside in an area of SF without an attendance area school? Has the district updated its maps so that folks in the Haight Ashbury now have an "attendance area school" for whatever that is worth? Our neighborhood school is actually a Chinese Immersion school, and then across the Panhandle we have an Alternative School, so that can't be our attendance area school.

    ReplyDelete
  4. Speaking as someone who lives in an assignment area that has no school now, and went through the lottery in one of its early years, we were told that our first non-alternative school (that is, first school with the designation of neighborhood and not alternative) that was on our list would be considered our neighborhood school. I don't know if that was true or is true now, but it may have helped: we got our first non-alternative designation.

    ReplyDelete
  5. Vicki,

    Could you tell us how we would find out how our attendance area is defined?

    "
    In the technical description of the SAS it states that for attendance area schools, attendance area kids within the applicant pool will be selected as long as their demographic profile increases diversity. Further it states, that once attendance area kids no longer increase diversity and all kids within the applicant pool are considered, if there are multiple kids within a selected demographic profile than the kid who lives within the attendance area will be selected for assignment. "

    Could you please explain this statement? In particular, the above comment states that attendance area kids will be selected as long as their demographic profile increases diversity.

    What if a kid is not defined as diverse? Do they go ahead of or behind kids from outside the attendance area?

    Also, I think there is some confusion about this statement:

    "as long as there demographic profile increases diversity."

    What about a kid that doesn't fill any of the diversity criterion applying to a school outside their attendance area?

    In other words, what happens when a kid that doesn't meet any of the diversity criterion applies to a school outside their attendance area like West Portal, Robert Louis Stevenson, or Sunset?

    Would they not go behind all children in the attendance area as well as behind all "diverse" children outside the attendance area?

    Thank you for any information you could provide.

    ReplyDelete
  6. This makes my brain hurt. I think I'll go private.

    ReplyDelete
  7. any of the diversity "criteria"

    ReplyDelete
  8. To 1:25 Its really too difficult to try and write out the lottery system algorithm. But yes, diversity (as defined by the SAS) trumps attendance area. Its not a slam dunk but it does give you a preference to be in the attendance area. Try making it to one of our Library Events or go to the SFUSD school Fair. Jeff Filling-Selk, the analyst who helped design the current SAS, usually gives a presentation on the SAS at this event. I haven't heard definitively if he will be there this year but I assume so. Also, read the technical description of the SAS.

    ReplyDelete
  9. To 1:09. 1:24 is correct.

    The first non-alternative school (that is, first school with the designation of neighborhood and not alternative) on your list would be considered your neighborhood school. - thanks 1:24 for your verbiage.

    ReplyDelete
  10. To 1:09 and 1:24. Woops. I spoke too quickly. Let me make a slight correction.

    You wrote
    "(that is, first school with the designation of neighborhood and not alternative)"

    It should say
    "(that is, first school with the designation of attendance area school and not designation of non-attendance area school):

    All alternative schools are non-attendance area schools but there are non-attendance area schools that are not called alternative schools.

    I know this is going to open up alot of questions about what is an alternative school so let me just say strike it from your vocabulary. It is a reference to a previous Student Assignment System.

    ReplyDelete
  11. So are all immersion programs "non-attendance area schools"? How about other language programs (JBBP or Russian FLES)?

    Is there a definitive list of "non-attendance area schools"? (The "alternative schools" are listed specially on the schools map, with their own colors and everything.)

    ReplyDelete
  12. Vicki,

    1:25 here.

    I think my question was pretty straight.

    Why aren't we clearly told how the algorithm works? I don't think it is beyond us. I should have to attend a meeting to find out.

    There has been a lot of complaining about parents opting out of the public school system and sending their kids to private school. Clearly, this group is almost completely from the group that doesn't fulfill any of the diversity factors. If the school board is sincere about encouraging this group to consider public schools, it also needs to fully disclose how the algorithm works for a non-diverse child.

    Also, why is it so hard to define the attendance area?

    Why isn't there online information to describe attendance area? Is there and I have just missed it?

    ReplyDelete
  13. correction:
    I should attend a meeting to find out?

    ReplyDelete
  14. To 3:11 & 3:22 You are absolutely correct. My apologies.

    For a complete disclosure of the current algorithm go to
    http://portal.sfusd.edu/data/epc/DI_Handout_Combo.pdf.

    To say, however, that just because your child does not have one of the characteristic that is part of the diversity index does not mean that your child would not add diversity to the school. The goal is a 50% mix of kids with the characteristic and those without. For example, more than 50% of the kids in SF qualify for free and reduce lunch and are English Language Learners. Not having that characteristic is likely to add to the diversity of the school.

    ReplyDelete
  15. Correction

    "Not having that characteristic is likely to add to the diversity of the school."

    "Not having that characteristic CAN add to the diversity of the school."

    But again going to

    http://portal.sfusd.edu/data/epc/DI_Handout_Combo.pdf

    will fully disclose how the algorithm uses diversity factors and attendance area.

    ReplyDelete
  16. To 3:10

    Depending on what map you are looking at the attendance area schools have one color and non-attendance area schools have another color.

    However, non-attendance area schools are:
    Claire-Lillienthal Alternative
    Yick Wo Alternative
    Argonne Alternative
    New Traditions Alternative
    Lawton Alternative
    Alice Fong Yu Alternative
    Sunset
    Diane Feinstein
    Lakeshore Alternative
    Clarendon Alternative
    Rooftop Alternative
    SF Community Alternative
    Harvey Milk Civil Rights Academy
    Buena Vista Alternative
    Willie Brown College Prep Academy
    Chinese Immersion School @ DeAvilla

    I think I got them all. All other schools have an attendance area.

    ReplyDelete
  17. This is an excellent example of why we should all RUN, not walk, to the PPS web site at http://www.ppssf.org and JOIN. Vicki and the other staffers there aren't volunteers, they're paid staff of an organization that exists solely to help us parents navigate the public school system. Having someone like Vicki figure out all this stuff and post is is gold, and it's something we should all be paying for. You can join PPS without having kids in public school. You can join PPS without have kids, for that matter. It's as little as $10 (but lots more if you can afford it). These women do astounding, important work and without them many of us parents would be lost. Join and help them keep others from being lost.

    ReplyDelete
  18. Thanks for that list, Vicki.

    Does that mean that someone who lives in the attendance area for West Portal will have a (slight) advantage for the Cantonese immersion program there? And also that someone who lives in the old DeAvila attendance area doesn't have an advantage for the Cantonese program there?

    Also, I don't imagine that anyone here qualifies, but do the bilingual programs for immigrants fall into the attendance area advantage?

    I find it pretty impressive that this system has been in place since, what, the 2002-03 school year, but these kinds of confusion are still so present. And the system is changing next year. When will that documentation be updated?

    ReplyDelete
  19. Thanks Beth. Yes we really could use the support and everyones membership.

    Yes, Frank that is what it means. Someone asked Jeff this question last year at the school fair and he answered affirmatively. Good question regarding the Bilingual programs but your child would need to be assessed as an English Language learner to qualify. After that my guess is that the algorithm would apply. Its probably similar to having a child who requires Special Education. I get abit shaky on my knowledge when it comes to ELL programs and Special Education programs.

    ReplyDelete
  20. Same question as Frank: does an immersion program in a school still give preference to "attendance area" applicants?

    ReplyDelete
  21. It looks like "yes." An immersion program in an attendance-area school will give preference to students in that attendance area. (So West Portal, yes, but DeAvila and Alice Fong Yu, no.)

    ReplyDelete
  22. Thanks for that doc, Vicki (http://portal.sfusd.edu/data/epc/DI_Handout_Combo.pdf). I'd never seen that before. It's complicated and it reads like two computer programmers talking to each other, but I understand more what we face.

    Essentially, the computer puts every incoming kindergartner into one of 16 pools. (Each pool is defined by the 4 diversity factors.) Then, after pre-assigning siblings, it calculates a "Composite Diversity Index" (CDI) with the students in the class. Then, it recalculates the CDI for each of the 16 pools and determines which pool would push the CDI closer to 0.50, and picks a student out of that pool. And then it starts again.

    I would imagine that there would be some advantage to being in a smaller pool, because if your pool is picked, you have a better chance to actually get a tentative assignment. So there's a balance between being in a small pool and being in a more-likely-to-be-picked pool.

    I guess that the attendance area advantage happens in that after the pre-assignment, only kids in the attendance area are evaluated. The attendance advantage ends when all of the populated pools would move the CDI away from 0.50. When that happens, the process is done with all the kids until there are no more spots.

    It actually makes "some" sense now.

    Something that's frustrating is that, while the district publishes racial/ethnic breakdown for each school and for each class in each school, the CDIs are top secret. If I knew the CDI for a particular school, I could calculate the likelihood of my family contributing to the diversity of the school (estimating from the sibling pre-assignment).

    It would be nice also to see the breakdown of the 16 Round I kindergarten pools for, say, West Portal GE, Clarendon GE, Rooftop Alt and Lawton Alt. Just how unbalanced are the applicant pools for these high-demand schools?

    ReplyDelete
  23. I don't see the information on how AAS and SES are defined.

    Academic Achievement Score (AAS) Socioeconomic Status (SES)

    How are these determined?

    ReplyDelete
  24. Frank, I also looked at the algorithm.

    There are four criteria:

    Home Language
    Extreme Poverty
    Academic Achievement
    Socioeconomic Status

    You get a one or zero for each of these.

    From what I am reading, the program "bins" applicants in the following way:

    0000
    0001
    0010
    ....
    1111

    across the four diversity criteria.

    English speakers with an advanced degree, and good socioeconomic status fall into ONE OF SIXTEEN bins.

    The program tries to pick equally across the sixteen bins.

    So, roughly speaking, as long as applicants in the other categories apply, the chances of getting into a school for the above group is about 7%.

    (I'm ignoring any advantage gained from being in the attendance area.)

    Am I wrong?

    ReplyDelete
  25. Note that I am assuming that the applicants across 0000 to 1111 apply equally across the spectrum. That's not true, of course. However, it is true that in all the "trophy" schools, a disproportionate number of affluent, advanced degree, english speakers apply.

    ReplyDelete
  26. Parent level of education DOES NOT affect placement.

    Poverty and socio-economic status are determined by:
    Does your family live in Public Housing? YES NO
    Does your family participate in Cal Works? YES NO
    Does your family receive Section 8 assistance? YES NO
    Is your family eligible for free/reduced lunch? YES NO

    Academic Achievement is determined by:
    Current or last school (including pre-school)

    Anybody know when the 2010-2011 Enrollment Guides will be available?

    ReplyDelete
  27. What, then, is the meaning of "Academic Acheivement?" referred to in:

    http://portal.sfusd.edu/data/epc/
    DI_Handout_Combo

    (The document that Vicki has pointed us toward as a reference.)

    ReplyDelete
  28. Can I ask the PPS folks if this new assignment system is going to apply to middle school applicants starting in the 2011-12 academic year? If so, how does academic achievement kick in. Is it the academic achievement of your child's elementary school or your child's actual STAR results that will count?

    ReplyDelete
  29. 8:26 PM

    I don't think the assignment system that Vicki is describing is the "new" system. I think it is the current system.

    Vicki, correct me if I am wrong.

    ReplyDelete
  30. PS.

    It would help if the document referenced had a date and title on it.

    ReplyDelete
  31. For K, Academic Achievement is determined by if the child went to pre-school or not. Meaning those kids that went to pre-K will be better prepared for K.

    ReplyDelete
  32. 8:20 PM:

    So my statement from 6:58 still stands accept that Academic Achievement is defined as preschool or not, rather than parents with advanced degree or not.

    The document does not appear to define Academic Achievement, so I think I can be excused for the mistake. Most of us don't think of sending our kids to preschool as "Academic Achievement."

    ReplyDelete
  33. oh you're right. didn't mean to call you out. i agree w/what you said.

    ReplyDelete
  34. It is interesting that poor/low SES get two points of diversity. I bet they are heavily correlated (i.e. very few applicants are extremely poor and not low-SES).

    ReplyDelete
  35. " Then, it recalculates the CDI for each of the 16 pools and determines which pool would push the CDI closer to 0.50, and picks a student out of that pool. And then it starts again."

    It'd seem to me that setting 0.5 as the target CDI isn't optimal: that the goal should be whatever the average is for the school district.

    ReplyDelete
  36. With regard to free lunch qualification, it's pretty close to 0.5: 54% of our SFUSD intake qualifies.

    ReplyDelete
  37. "
    It is interesting that poor/low SES get two points of diversity. I bet they are heavily correlated (i.e. very few applicants are extremely poor and not low-SES).

    October 29, 2009 10:00 PM
    "

    6:58 from last night here.

    I noticed that too. I was wondering if someone else would pick that up.

    Thanks.

    ReplyDelete
  38. "
    It'd seem to me that setting 0.5 as the target CDI isn't optimal: that the goal should be whatever the average is for the school district.
    "

    Yes, thank you for pointing this out.

    The program actually seems pretty clunky to me. One thing it could do is try to adapt to the applicant group, thus sparing us somewhat of the Round II and Ten Day Count Ordeal. (As well as saving some people their lost private school deposit.)

    ReplyDelete
  39. To 4:37. Yes, according to Jeff at last years school fair an immersion school that is an attendance area school gives preference to "attendance" area applicants. So kids within the Flynn attendance area would have slight preference while Buena Vista gives no preference.

    ReplyDelete
  40. To Frank at 6:15pm

    Yes, that's it. Attendance area comes into play again, even after all city kids are considered. When the SAS goes to a particular pool and there are multiple kids in the pool it will give the assignment to the kid who lives in the attendance area.

    Remember that it would be nearly impossible to know what the CDI is because it is made up of siblings at first and then siblings plus the selected applicant. So what pool it chooses from changes. I tried to do a simulation using a school I thought I knew the demographic profile of and found thre were just so many unknowns that it is almost impossible to make any prediction with any accuracy.

    Bottom line - Supply and demand is the biggest determinants and parents are best off looking at those numbers rather then trying to figure out how the diversity index will help or hurt them.

    ReplyDelete
  41. This conversation (which also makes my brain hurt) is a good illustration of why the current system is not working and why we are redesigning it.

    If I could offer my personal opinion here, I would like to say that I agree with Vicki 100 percent -- it is far more productive to spend your time looking at the supply/demand numbers for the schools you are interested in than it is to try to figure out how the diversity index will apply to you.

    There is no documentation for the new system because we haven't decided on what the new system will be yet. The plan is to have a new policy approved by the Board as of March 2010, for implementation in the Fall of 2010. Implementation would presumably include creating documentation/explanations to help parents understand the new system!

    To the person who asked whether it would apply to students entering middle school in the fall of 2011, my presumption is yes, that it will be implemented throughout the system.

    ReplyDelete
  42. In reference to 8:20pm. The Four socio-economic indicators are described on page 20 of last years enrollment guide. This years enrollment guide as well as this years application will be distributed at the school fair on Nov. 7th. After that it will be available on line.

    Below is the description of the four indicators for Kindergarten.

    • Extreme Poverty: Does the student live in public housing? Is the student a foster youth? Does the family participate in a homeless program?

    • Socioeconomic Status: Does the
    student participate in any of the following programs: free/reduced lunch, CalWORKS, and/or Section 8 housing?

    • Home Language: Is English the student’s home language? This is determined by the answers to the Home Language Survey on the application form. (Vicki Note: This year ELL students will be determined by an English assessment prior to completion of the application)

    • Academic Achievement Status:
    - Incoming Kindergartners: Did the
    student attend preschool?

    ReplyDelete
  43. To 8:23

    What, then, is the meaning of "Academic Acheivement?"

    Did your child attend preschool or not.

    ReplyDelete
  44. To 8:26 and to reiterate what Rachel Norton posted - what we are describing here is the current Student Assignment System. One of the goals of the current SAS was to achieve socio-economic diversity through a choice system. The hope, I believe, was people would self diversify. After all it worked and is still working to a large degree with the old Alternative schools such as Rooftop and Lakeshore. Unfortunately, these schools provide busing from all over the city whereas most of our attendance area schools do not.

    In actuality, the current SAS did not achieve these goals of socio-economic diversity and at the same time is difficult to understand.

    What the new Student Assignment System will look like has not yet been determined. Several models are on the table and will be vetted with the community. The PAC and PPS-SF will be holding a number of input sessions with all communities in San Francisco. The feedback will then be presented to the board and will serve as one of the considerations the board will need to contend with.

    ReplyDelete
  45. To 8:30

    " PS.

    It would help if the document referenced had a date and title on it."

    I agree - its generally referred to as "The technical description of the Student Assignment System."

    ReplyDelete
  46. To 10pm:

    Yes I have often noted that. How would you be Extremely Poor and not also be of low Socio-economic status. The result, in my opinion, is that it would reduce the number of actual buckets.

    Also, keep in mind that kids who are Extremely Poor are the least likely to participate in the lottery system at all.

    In reality we have two student assignment systems. The first is The people who choose and the schools they choose the second is the people who don't participate and the schools that don't get chosen.

    My guess is we have gotten very technical and theoretical here and lost the people who were most interested in the simple question of attendance area impact. (Is anything simple?)

    For those of you who want to discuss this indepth let me suggest you attend Jeff's session at the School Fair if he is giving it again. As soon as we know what the workshop schedules are we will post them on our website. Go to www.ppssf.org and while you are at it become a member - its only $5 for an individual $10 for a family. Keep our network alive active- Parents helping parents.

    ReplyDelete
  47. "Yes I have often noted that. How would you be Extremely Poor and not also be of low Socio-economic status. The result, in my opinion, is that it would reduce the number of actual buckets."

    No, but you can be of low socio-economic status and not be extremely poor. It functions by distinguishing the poor from the very poor.

    ReplyDelete
  48. To the person who asked whether it would apply to students entering middle school in the fall of 2011, my presumption is yes, that it will be implemented throughout the system.

    Hi Rachel--would this include high school? I had thought from the earlier descriptions it would only be K-8. With so few high schools, with so many different programs, why limit the kids?

    ReplyDelete
  49. Vicki and Rachel,

    Thank you for the information.

    Clearly, the people running the application lottery are trying to balance many factors.

    Still, the process seems to be bending over backward to get socio-economically disadvantaged lottery non-participant families enrolled in reasonable schools. This is at the expense of AAS, non socioeconomically disadvantaged, English speaking, non EP families. Let's call them the middle class. The middle class must surely comprise more than 7% of the total applicant pool.

    Parents for Public Schools have been selling the idea for years that "the Public Schools are Good."
    That is true. Many are. But middle class families have a very high probability of not getting access to one of these good schools: a more than a 93% chance of failure.

    I don't disagree with the attempt to help disadvantaged children, but it has been clear for a number of years that many poor families don't participate at all in the system. Why flog a dead horse? The lottery is not going to fix this problem.

    The middle class participants in the lottery are put through several rounds of virtually hopeless applications so that the non-participants might eventually sign their kids up for a decent school. Meanwhile, the middle class are almost apoplectic at the process and have mostly voted with their feet.

    The few remaining middle class families are either in the lucky 7%, or have played slightly better odds on the ten day count (largely because of all the families that dropped out) or managed to apply to a school such as Alamo that doesn't have as many poor, English language learner, no preschool applicants.

    Unfortunately, most us living in the SE and are not that keen on schlepping to Alamo or Yick Wo.

    It is true. The process and the algorithm make my head hurt too. I hope we can come up with a better process and algorithm.

    ReplyDelete
  50. There's another huge practical problem in including high schools in the system. The high school map in current existence has a gynormous hole in the middle of the city for what had been the zone for the old McAteer. With apologies to the elementary school applicants in the Haight, this is not a small hole that impacts one neighborhood, this hole encompasses all of Twin Peaks/Midtown Terrace, Diamond Heights, Miraloma, Forest Hill, Golden Gate Heights, Noe Valley, and parts of the Castro, Inner Sunset, the Haight (you guys screwed again!), Glen Park and various bits of other areas including what appears to be a small sliver of Bernal! I don't know how SFUSD could apply it to high school, unless those maps were radically changed!

    ReplyDelete
  51. Vicki at PPS-SF said: "In actuality, the current SAS did not achieve these goals of socio-economic diversity and at the same time is difficult to understand."

    I agree that the current SAS is difficult to understand. But did it fail to achieve the socio-economic diversity goals? It's weird that the SFUSD website lists the per-class racial/ethnic breakdown, but doesn't list the per-class diversity factors. We can find out what percentage of the Lilienthal 2nd grade is Latino, but not which percentage is a high academic achiever. And how much of that failure is due to the fact that the SAS only works for Round I and not waitpools or Round II or 10-day Count?

    Vicki at PPS-SF said: "In reality we have two student assignment systems. The first is The people who choose and the schools they choose the second is the people who don't participate and the schools that don't get chosen."

    And this is the root of the problem. Will the new SAS really address this?

    One of the amazing things about the current system is that it got the Choosers to choose Unchosen schools (Miraloma being the most famous example, but language and other magnet programs, too). By turning Unchosen schools into Chosen schools, the district as a whole is better (best urban district in CA, top 5 urban districts in the US, etc).

    Now what to do to turn the Non-Chooser parents into Choosers? One is to clarify and simplify the process, but a big one is outreach. What is the District doing today to inform these parents that the deadline for kindergartners is in January? Do we even know what the deadline is yet? How about for at-risk families in good schools who have younger siblings coming? (I'm always amazed that there are Siblings on the waitpools at Rooftop.) Is there a District employee responsible for this for kindergartners enrolling for the 2010 school year?

    I hope the new SAS doesn't throw the baby out with the bathwater.

    ReplyDelete
  52. Regarding this question:
    ----------
    Yes I have often noted that. How would you be Extremely Poor and not also be of low Socio-economic status. The result, in my opinion, is that it would reduce the number of actual buckets.
    ---

    They are indeed two different buckets - Extremely Poor = qualifies for public housing while low socioeconomic = qualifies for free and reduced lunch. I don't know the specifics, but each has a different threshold. Fewer people are extremely poor, more are low socioeconomic.

    ReplyDelete
  53. I don't think the middle class has a 93% chance of failure. The chance of failure is very high if they continue to apply to the top 11 schools (of which Alamo is one, Yick Wo is not). There are many other schools with good API scores, good relative-to-demographic API scores, good magnet programs, solid teaching, committed principals, engaged families. That's why we are one of the top urban districts--because of breadth of good stuff happening, not just at a few schools. Marcia Brady and June have reviewed just a few of these others. Apply to these with a balanced list, and your chance of getting a good school goes way, way up. The trick is to avoid the uber-popular (unless you are willing to do the 10-day-count thing, which is fine if you are up for that).

    ReplyDelete
  54. 12:38 PM:

    I've already stated my assumptions above.

    I assumed that the 16 applicant pools each are full enough so that when the algorithm pulls equally from each pool, it comes up with a uniform distribution of kids across the diversity spectrum, 0000 ... 1111.

    If that is the case, each "pool" forms 1/16th of the total students in each class.

    Of course, it is not always the case that each applicant pool is full enough for that to occur. On average, however, for most schools, "trophy" or not, that is what happens.

    The middle class, who are in a single bin, form only 1/16th, or approximately 7% of the successful applicants.

    That is how Round I works, ignoring immersion applicants and a minor adjustment for alternative and non-attendance schools.

    I would also add that families are put through this process three times; for kindergarten, middle school and high school.

    For middle school and high school, the odds become even more daughting with the addition of yet another statistic, the API of the elementary school. That divides the bins into 32. An astute observer will note that it creates a disincentive to improve the API elementary schools.

    ReplyDelete
  55. correction:

    . . . API of elementary schools.

    ReplyDelete
  56. Actually, I think that academic achievement for grades other than kindergarten is determined by the API of the sending school, not the scores of the student. Thus there are still only four criteria for determining diversity, not five. So attending a school with low API scores gives students a leg up in being assigned to middle and high schools that are more attractive by middle class families.

    ReplyDelete
  57. 1:57 PM

    Perhaps you misread my statement.

    That is exactly what I said: API of the sending school. For middle school applications, the sending school is the elementary school. A very organized and determined PTA like the one at Miraloma will raise the API of the entire school. In doing so, they lower their chance of their child getting into their chosen middle or high school.

    "Thus there are still only four criteria for determining diversity, not five."

    Wrong. Direct quote from Vicki's document:

    "In SFUSD’s implementation of the Diversity Index, there are five characteristics
    for non-kindergarten applicants and four characteristics for kindergarten applicants:
    Academic Achievement Score (AAS), Socioeconomic Status (SES), Extreme Poverty
    (EP), Home Language (HL) and – for non-kindergarten applicants – API Rank of
    Sending School (ARSS). Each characteristic chosen by SFUSD is used in a binary form,
    so an applicant will have either a “1” or “0” for each of the four or five characteristics
    used. Because there are five binary characteristics, there are 32 possible profiles for nonkindergarten
    applicants."

    Note: FIVE characteristics leading to 32 possible profiles for nonkindergarten applicants.

    So if you are middle class and your child goes to a school with a high API, when you apply to middle school or high school, your chances of success in the lottery are less than 1/32 or 3%.

    ReplyDelete
  58. Thanks, 2:21, I stand corrected.

    For me, parsing the diversity index algorithm is a waste of time. It's kind of like childbirth: Understanding its minutia seems to be immensely important to many who have never gone through it (i.e., prospective kindergarten parents) and completely uninteresting to everyone else (i.e., those whose children are already enrolled in a public school, including those who were terribly frustrated and fixated on the process beforehand). I'm sure there are exceptions, but I have never heard any parents discussing the diversity index in relation to middle and high school applications.

    ReplyDelete
  59. 1:30 pm:

    I assumed that the 16 applicant pools each are full enough so that when the algorithm pulls equally from each pool,

    I doubt that all these "pools" fill up in the applicant process in Round 1. Or else Clarendon would look a lot more diverse along SES lines than it does.

    Assuming that not-poor, not-very-poor, English-speaking, preschool-attending families constitute exactly 1/16 of an applicant pool at a given school, they would all get in, right? 7% is a little low for the district, but if we were spread across the district in equal proportions at each school, it's probably not that far off. However, we cluster in huge numbers at Clarendon, Miraloma, Alamo, etc. and then we run into trouble. Ultimately the alogorithm abandons the diversity principle anyway and just takes whomever, using neighborhood as a tie-breaker and eventually using rank choice as a tie-breaker, and finally just going random. That's how Clarendon ends up with so many not-poor kids. But by then it is the same not-poor etc. families competing against each other in a huge pool--that 1/16th set but it is probably more than half of the applicants (just guessing here). That's a huge number to overcome.

    We would all be better off applying to schools with fewer applicants overall--and fewer applicants in our "pool." But I agree with Rachel that the best thing is simply to avoid the most super-popular schools.

    ReplyDelete
  60. "
    I'm sure there are exceptions, but I have never heard any parents discussing the diversity index in relation to middle and high school applications.

    October 30, 2009 2:39 PM
    "

    Well, I guess I know of a few specific exceptions: parents who were accepted to Clarendon who decided to go private because of the low chance of successfully navigating middle and high school process. That was also a factor for us, even though we went 0/7.

    I think your post is kind of smug.

    Statistically speaking, the sited document demonstrates that the lottery process is highly skewed against the middle class for kindergarten, middle school and high school.

    No words or glib statements can hide that fact.

    ReplyDelete
  61. 2:54:

    What your telling me is that the middle class are slightly better of applying to schools like Clarendon and Alamo, rather than schools like West Portal and Alvarado.

    Amongst oversubscribed schools, the last thing you want to do is apply to a school like Sunset, West Portal or Alvarado that likely fill up all their applicant pools.

    And sorry for parsing the diversity index! Actually, I am no longer an interested parent applicant to the public schools. I consider this method of enrolling students to be a betrayal of public trust. I've lost interest in wanting to deal with such a system.

    ReplyDelete
  62. For High School, it appears that Mission High and John O'Connel have become the de facto, alledged, neighborhood schools.

    ReplyDelete
  63. Yes the diversity index is completely skewed against middle income families that have sent their kids to pre-school. The diversity index trumps all, so unless you live near a school like John Muir which has more space available then students, you aren't getting into your neighborhood school. I can't imagine that there are more than a handful of schools with under-enrollment in kindergarten, which means that it is a garuntee that families interested in their neighborhood school stand a good chance of not getting into that school, meanwhile families bringing the same type of diversity as the family from the neighborhood, bump the negibhorhood family. Sorry, but that isn't okedokee.

    ReplyDelete
  64. 3:12: High school can be tough. However, most families, actually all families that I know, got one of their choices for middle school, and almost all got their top choice. Middle school lottery has simply not been a barrier. (This may be in part because of low sibling preference.) It doesn't make sense that a family would go to private school out of concern of failing in the lottery at that level. One would at least want to try, I would think--applying to public school is free.

    3:25, I think Clarendon and Alamo as well as Alvarado and West Portal are all super-over-subbed, and I would strongly urge parents not to put any of them down as #1 or #2 unless they are prepared to go through the waitpool process. Far too many middle class parents apply to them--forget the diversity index for a moment--you will end up in a large pool of people just like you. It will be sheer luck if your kid's name is picked as one of 950 equally not-poor kids in that pool or whatever the mix is.

    It sounds like the new options for consideration may alleviate this pressure somewhat by giving a leg up to kids from very low API neighborhoods (guessing there won't be huge numbers wanting to trek across town) and then neighborhood preference folks.

    What part do you feel is a betrayal of public trust? The byzantine algorithm that is so hard to understand? Or giving a leg up to poor folks? I don't think that last part actually has a huge impact on Clarendon, based on who actually applies.

    3:37 I don't believe families who are outside the neighborhood actually do bump neighborhood families with the same diversity profile; the opposite, in fact. The algorithm gives tie-breaking preference to neighbors after seeking diversity from all pools. The problem really is that too many people of the same diversity profile apply to the same small numbers of schools. There are ways to avoid John Muir if you want, but it usually involves openness to schools that are not geting thousands of applicants. There are actually, statistically, many schools that are in the upper end of achievement that are not getting thousands of apps.

    I'm really not trying to be smug about it, just stating the facts. Objectively, there are schools with various combinations of good test scores, magnet programs, and community support such as PTAs that are not among the 11 schools that are way, way over-subbed. I think that is good for prospective parents to know--that there options in between Clarendon and John Muir. Read the reviews going up on this site and you will learn about some of them. Go to the School Fair on November 7 and learn about more. It's a sucky, byzantine process, yes, but it is possible to get a decent school in the lottery if you look in the right places.

    ReplyDelete
  65. "Well, I guess I know of a few specific exceptions: parents who were accepted to Clarendon who decided to go private because of the low chance of successfully navigating middle and high school process. That was also a factor for us, even though we went 0/7."

    Oh, my mistake . . . I meant families who are actually applying for middle and high schools. And what does attending Clarendon have to do with avoiding public middle and high schools?

    ReplyDelete
  66. 4:13
    Actually the bumping does happen. It happened to our family. One white middle-income, English-speaking only, pre-schooled family from the neighborhood was bumped by a family with the same characteristics from out of the neighborhood. And the end result is that most of the families that were enrolled are middle class white - mostly from out of the neighborhood. Go figure! Don't defend the algorithim or try to make sense of it. I have been through this process three times now, and each time with a different result.

    ReplyDelete
  67. I hate to say it, but I am actually glad to be going through the assignment process this year. Next year, I do fear, will be riddled with all kinds of errors.

    ReplyDelete
  68. To 10:59 AM,

    Part of the redesign plan will be to clean up the 30 year old school boundary lines.

    ReplyDelete
  69. 10:03 AM

    Thank you Vicki for telling us about the boundary lines.

    Some of us in South Bernal are in the 94112 (Excelsior) zip code. It would be helpful to have us put in the Bernal school area.

    ReplyDelete
  70. I find this very disheartening. We pray and hope to get get into an elementary school that has decent test scores to find out that its going to hurt us in middle school. No wonder folks go private if they can afford it.

    ReplyDelete
  71. 2:28, middle school just hasn't been a huge problem, even in Round 1. We and ALL of our friends got their first choice when we went through it. Then one family forgot to enroll in the school (oops) but they were able to to get a slot back eventually. Remember that middle school has very little in the way of sibling preference because it is only three years. Also, parents are just more relaxed by the time middle school happens. At least, that is my observation.

    ReplyDelete
  72. At our elementary school for the past two years, every kid got their first choice middle school. And all the middle schoolers I know all think they are in the perfect school for them (even though these are all different public middle schools.)

    The enrollment dynamic is a completely different than for elementary.

    ReplyDelete
  73. All this talk of neighborhood preference and Clarendon. I didn't think Clarendon had a "neighborhood". It is an alternative school, no?

    ReplyDelete
  74. "Statistically speaking, the sited [sic] document demonstrates that the lottery process is highly skewed against the middle class for kindergarten, middle school and high school."

    There's no statistics in the document to support that conclusion. The numbers are from a hypothetical simulation.

    ReplyDelete
  75. "I doubt that all these "pools" fill up in the applicant process in Round 1. Or else Clarendon would look a lot more diverse along SES lines than it does. "

    IIRC, Clarendon has a low %age free/reduced school lunch students (19%, versus 45% district average); same with ELLs.

    The reason it's difficult for middle/upper middle class kids to get into Clarendon is because there's so many other middle class applicants, not the unwashed poor or foreigners hogging all the spots.

    ReplyDelete
  76. Question about attendence areas. If by the map my local area school is for example Daniel Webster, though by physical distance I'm closer to Starr King, whose Mandarin program I'm really interested in, am I considered out of area for Starr king? Does actual proximity to the school count at all?

    ReplyDelete
  77. "Question about attendence areas. If by the map my local area school is for example Daniel Webster, though by physical distance I'm closer to Starr King, whose Mandarin program I'm really interested in, am I considered out of area for Starr king?"

    Yes.

    "Does actual proximity to the school count at all?"

    No. Some attendance areas aren't contiguous: IIRC some of Yick Wo's attendance area is in HP/BV.

    However, *in theory* immersion programs shouldn't have an attendance area; they should work the same as the atalternative area schools. Now, last year the theory and the practice didn't match in the actual code, and the immersion programs actually did have attendance areas. Check w/PPSSF as to whether the district has changed the code to reconcile policy with the algorithm. I don't know how Fairmount, which is an attendance area school which is wholly immersion with no GE program, would work under the policy.

    ReplyDelete