Sunday, November 29, 2009

Hot topic: How to rank your 7 schools

This from an SF K Files reader:
I am hearing mixed messages about the importance of the order of the 7 school choices on the SFUSD application and wonder if readers could help clarify. In a few comments on this blog people who seem like they know what they're talking about have stated that the ranking is irrelevant, and yet Rachel Norton posted on her blog that " Of the 947 families who did not receive any of their Round I choices last year, almost 800 listed one of these high demand schools as their first or second choice" which seems to imply that the order does matter. What gives?

48 comments:

  1. I'll just keep posting this over and over, I guess. Some people who sound knowledgeable do indeed say that rank order does not matter, because that was the commmon wisdom until last year when Vicki of PPS sat down with the person who wrote the code and found out for sure otherwise. But the old understanding persists and has to be countered again and again on this blog! Here is the updated version:

    Rank order can matter way down the line as a tie-breaker. The computer first looks for diversity with the assignment area, then outside the assignment area. When there is no diversity left to be found--for example, at Clarendon where so many of the applicants have the same (not-poor) diversity profile, then rank order is used as a tie-breaker, as is neighborhood in non-alternative schools.

    Therefore, in a large (huge, enormous) pool like Clarendon, it is likely that only people who put it #1 or #2 will be chosen if they don't add diversity to Clarendon's current profile. So if you add diversity--for example, many here would add diversity to the amazing school E.R. Taylor--then rank order won't matter, since diversity matters more for you. But if you are applying to a school whose diversity profile looks like you, and the applicant pool is very large, then putting it below #2 is probably a waste of a spot unless you are using it as filler to go 0/7 and do the waitpools.

    On the other hand, if you list Clarendon at #1 and #2 then it is probably also a waste as it is a huge, huge longshot, so you'll probably be doing the waitpools anyway!

    My best advice in terms of rank order strategy is to list "bubble schools"--moderately popular schools that get more apps than spots, but not triple/quadruple/quintuple or more--I'm talking about schools like Harvey Milk and Sunnyside and Peabody; or up-and-coming schools like Paul Revere. List THESE schools #1 and #2. Rank order could make a difference for you as a tie-breaker as the last few slots are allocated in these schools that will surely fill up in Round 1, but not be virtually impossible if you play your strategy right.

    Hope that helps rather than confuses.

    ReplyDelete
  2. By "diversity", you just mean one's income, right, since ethnicity is not taken into consideration? And specifically, it's just whether you are low-income or not....

    ReplyDelete
  3. This is 2:33 again. Yes, 8:51, you are correct: diversity refers only to the socio-economic factors considered in the diversity index, such as whether you live in public housing. Race and ethnicity are NOT factors.

    Also, it is a binary system with regard to all these factors, not a sliding scale: you either are poor or not-poor along several indices--people who are working class, middle class or rich are all treated the same if they don't qualify as "poor." When you apply to a school like Clarendon, you are mostly competing against people who are not-poor--that group is very large. The computer gobbles up those who do add diversity and then applies tie-breakers to the rest; and even then there is such a large group that it is sheer longshot luck if you get one of those remaining spots. That's why, if getting a school in Round 1 is a top priority, you're better off listing one of those moderately popular schools as #1 and #2, as mentioned above.

    ReplyDelete
  4. Thanks 2:33 for clarifying the assignment process and also giving your take on how that translates into a strategy for round 1. Much appreciated!

    ReplyDelete
  5. What 2:33 pm said about rank order is correct: it was thought up until last year that rank order didn't make a difference.

    So, you have a couple of different strategies, depending on your situation and tolerance for risk and uncertainty:

    1. [Safest] List no "trophy" schools on your application, focusing on the "bubble" schools e.g. E.R. Taylor, Milk, and Revere that have not yet become hugely oversubscribed in terms of first-place requests (you can get historical information on number of first place requests from the Enrollment section of SFUSD's website).

    2. List a popular but not trophy school in first place, e.g. McKinley or Fairmont or Monroe SI, and then fill out the rest with other non-trophy schools, using "bubble" schools as the safety option for your lower choices. Can work, but is more risky.

    3. List *one* or two trophy schools at the top, and then use same strategy as #2 above for the rtest. More risky; advise putting a "bubble" school in the #2 or #3 slot if you don't want to go 0/7. I have heard of people getting into trophy school who put them in second place (I'm one myself), but it is rare.

    4.[Riskiest] List more than 1-2 trophy schools. High chance of going 0/7. Probably a waste of the lower picks: only do this if you have a non-SFUSD Plan B (or C) and are willing to spend the summer in the waitpool.

    5. If the list of schools you really want is shorter than seven, trophy schools can be used to "fill" out the bottom slots of your application, as you're unlikely to get a place in those schools unless you're in a different diversity pool from most of the other applicants. This is advisable as those who go 0/7 are in a higher priority waitpool who list less than seven schools but go 0/6 or 0/4 etc.

    6. However, make sure logistics are going to be OK for all the schools on your application - you never can know what's going to happen.

    ReplyDelete
  6. Don't forget about Creative Arts Charter School (a free public school with it's own, non-bizarre lottery system). If it moves as planned to a better location, it has the potential to be a really popular school.

    ReplyDelete
  7. someone explain to me why the spam cant be deleted? isnt anyone monitoring this blog?

    ReplyDelete
  8. If rank doesn't matter (except for a tie), then why does SFUSD keep track of first place requests for each school?

    ReplyDelete
  9. Maybe to track the "probability" of getting your top choice or to track the desirability of a school?

    ReplyDelete
  10. 12:26,
    I think it's because it matters a whole lot what school you put as #1.
    One smart move, in my opinion, is to put as #1 a school that had a low-ish number of first choice requests last year. I think that's more important than total requests. Just a hunch, though.
    Then fill the list with your Miralomas so you can go the 0/7/waitlist route if you don't land a spot in the first round. Then at least you'll have a shot at getting that letter in March and be DONE.
    Worked for us, anyway!
    Good luck everyone--

    ReplyDelete
  11. "If rank doesn't matter (except for a tie), then why does SFUSD keep track of first place requests for each school?"

    Rank does matter. It also matters in choosing which school your kid is allocated to if they get into more than one school,

    e.g. let's say during the running of the algorithm, your kid is allocated slots at two schools - first one that is #4 on your list, and then one that is #2 on your list. The algorithm would then free up the slot that you ranked at #4.

    ReplyDelete
  12. 8:55 a.m. We followed this strategy, as suggested in a previous thread (perhaps you were the one who offered it). We are keeping our fingers crossed that it works.

    ReplyDelete
  13. It totally matters. Think of it this way. Take Alvarado for example. It had 476 total requests for 44 K seats (ok, less than that w/siblings we all know). But it only had 45 1st choice requests. Assuming most of those 476 requests are non-poor (b/c they turned the application in on time), and you are too (b/c you have a computer and are reading this blog), really just the 45 1st choice kids are duking it out for those spots. The lotto will only dip into the 2-7 kids to add diversity, but if it finds none, rank breaks the tie, and it is back to the 1st choice kids. So really you have a much better shot at Alvarado GE with 45 kids requesting 44 spots than the total 476 kids trying for 44 spots. Get it? Choose your number one carefully. Statistically if you do not get your number one or number two choice you will not get any of the seven. Lilienthal got 232 1st choice requests for 6 seats. Of course someone gets those seats, but your odds are low! Good luck everybody!

    ReplyDelete
  14. Lilienthal has 66 seats in GE, not 6.

    ReplyDelete
  15. But if some people end up with their #6 or #7, as people certainly have on this blog, how can it be true that it's #1 or 0/7?!

    If that is true, can I just say that it would make a heck of a lot more sense to take all those who don't get their #1, and make their #2 into a #1 for a second pass, changing all the #s accordingly, and going on down like that? Why should someone denied their #1 school then have to lose their #2 to people who chose that school as their #1? How can people who don't get their #1 effectively lose all the way down the line?

    Yes, you have to choose your #1 carefully, but even with the most careful choices, there's an element of randomness that it seems to me should not be punishable by 0/7. Or why have 7 slots at all? Why not have everyone submit 1 choice, and the losers go to whatever school is left with slots?

    Every time I think I "get" the lottery, something slips for me. My three advanced degrees are totally useless for this process.

    ReplyDelete
  16. 9:33, for every change in the algorithm, something is gained and something is lost. In a system of scarcity of "perceived to be good" seats, they can't make it so that everyone gets exactly what they want. It's an allocation system and someone will lose....UNLESS people start to spread out and choose seats beyond the scarce ones. The current system is designed to encourage people to do just that. If we were given only one or two choices, how many would just shoot the moon for Clarendon? A lot.

    So instead we are given the option of choosing 7--in fact we are encouraged by the "cohort" waitpool system to pick the max of 7--and that has encouraged people to look beyond the original dozen or so that were considered "acceptable." Now as many as 50 are in that category, so the strategy has actually worked.

    As the public schools have become more popular in recent years, however, formerly easy-to-get schools have entered the no-guarantee territory too, which has caused us all much angst as we worry about the best strategies for getting something we like! Used to be you could put Clarendon first and be guaranteed to get Miraloma or Sunnyside....not now. Of course, you can still feel pretty hopeful about getting Daniel Webster--gotta keep your eye on the up-and-coming schools.

    ReplyDelete
  17. 9:33 here. I am all about the up-and-coming schools. Not even one of the "notorious 11" are on my list. But I'm still confused. Supposing I put the most popular of my "non-11" schools at the top of my list, because I want to be in the tiebreakers. If I don't get that #1, do I then slide down the hill of doom into random placement? Or does having other up-and-coming schools at 2, 3, 4, and so on mean that I will likely end up with one of them? The latter would seem logical, but the explanations here don't support that logic.

    ReplyDelete
  18. "Supposing I put the most popular of my "non-11" schools at the top of my list, because I want to be in the tiebreakers. If I don't get that #1, do I then slide down the hill of doom into random placement?" "Or does having other up-and-coming schools at 2, 3, 4, and so on mean that I will likely end up with one of them?"

    I don't think you can easily estimate that from the data, given that stats for ranking below 1st place aren't available.

    "The latter would seem logical, but the explanations here don't support that logic."

    It All Depends. There were an unfortunate 5% of applicants who didn't list the trophy 11 as #1 or #2 and still went 0/7. Also, folks are more aware of the effect of ranking of the schools on probability of getting in, so I'd say that a lot of folks are going to be more strategic and the schools just below the trophy 11 will rise in the number of 1st place choices. So I'd expect a big jump in e.g. the number ranking McKinley or Sunnyside as a 1st place choice.

    Look at the ratio of applications to the school and the number of 1st choices for the school. Knock off 1/3 off the nominator and denominator for sibs. Take a look at the %age of free/reduced school lunches and ELLs to get an idea whether your kid is going to be in a different diversity pool to those prevalent in the school. By doing that. you'll have a better feel as to what your odds look like. See http://portal.sfusd.edu/data/epc/09-10%20round%201%20results/5%20year%20Demand%20Data.pdf for five-year data on # of applications.

    Now, if the school has a little bit of "buzz", you might see the number of 1st choice applications climb, which will throw off your calculations a bit. My impression is for the GE programs that these are a lot less volatile than total applications: e.g. Sunnyside saw its total applications rise from 62 in 2008 to 94 in 2009, but the number ranking it 1st dropped from 18 to 16. Similarly, applications to Flynn's GE program went from 56 to 94, a huge climb, but as a 1st choice it was unchanged at 8. McKinley went from 146 total applications to 203 (a 35% increase), but the number ranking it at 1st place went up by a smaller percentage - from 40 to 45 (a 12% increase).

    For immersion programs, though, given their popularity, I'd play it more safe, i.e. think about putting Marshall or Revere over the more popular Flynn or Monroe.

    Again, it all depends on your risk preference. Friends of mine got into Clarendon 3 weeks after Kindergarten started, so "shooting for the moon" can work with persistence.

    HTH.

    ReplyDelete
  19. I agree with 8:47's advice.

    If a school is truly under-subscribed, you will almost certainly get it; Rosa Parks JBBP is a great example of a little gem of a school community, with great extras including Japanese culture and language, that has gone under-subbed the last few years. It would be a great lower rank add-on if you are willing to go there but it is not your first choice. (I think it's a fine choice for #1 too--it's really a great community, especially for north side and Geary corridor commuters--I'm just using it here as an example of a safety school for the lottery). A great choice if you really don't want to go 0/7, and you like the school, but you are want to shoot for a more popular school with your #1 choice. In other words, the more you put a popular school as your #1 and/or #2, the more you should be put truly less popular schools as #3-7. Moderately popular probably won't work at #3-7. Unless you are are okay with 0/7 and waitpool of course.

    However, if your diversity profile differs greatly from the school as a whole and from the likely applicant pool, then it will more likely be a hit no matter where you rank it. Now that ER Taylor is getting buzz here, it will be interesting to see if anyone applies--it will be a good test of this strategy if so, as it is a VERY popular school but not among the middle/upper class set so far, so presumably it will be easy for people here to add diversity to that school in lottery terms.

    ReplyDelete
  20. That is correct; while they are "balancing," rank does not matter. If you are likely to balance a school at any point, no need to list it first.

    ReplyDelete
  21. In regards to 10:48 AM, if my child is allocated a slot at my #1 as well as my #7, is it really true that I am guaranteed to get my #1 ?

    I don't want to jeopardize my chances of getting my #1 (which is not a trophy school) by including "safe" choices in my seven. If I was the school district, I would give me the #7 and reserve the #1 for another parent that was more discerning in their choice of seven - that way I could claim a higher percent of parents who got one of their seven.

    In other words, I am inclined to list as #1 and #2 the only two schools that I really want and to fill my seven with trophy schools which would all be logistically acceptable (in the unlikely event that we got one). And, I would expect to go 0/7, but stand a very good change of getting my #1 in round 2. Does this strategy make sense ? What am I missing ? (I belong to what I expect is the largest pool : not poor, English only, pre-school yes)

    ReplyDelete
  22. Also, doesn't the system overlook first choices once it has filled that diversity quota? That would mean that Kid A might be overlooked, even though it's his first choice, and Kid B would be selected even if Kid B put the school farther down on his list because he matches the diversity combo sought by the computer.

    Then Kid A might get assigned a school that was actually Kid B's first choice because Kid A matches the diversity sought at that school. Then you have two families who didn't get their first choice. Am I understanding this right?

    But if you only put one or two 'real' choices, followed by 5 trophy schools, you're more likely to go 0/7 and then at least you have a chance to wait it out in a better cohort.

    No matter how you look at it, it's a wasteful process for many, many families, but most do get something they can live with in the end. I do believe in helping kids rise above their circumstances, but obviously the social engineering comes at a cost to many families in SF.

    ReplyDelete
  23. "Also, doesn't the system overlook first choices once it has filled that diversity quota?"

    There's no diversity quota. The algorithm works in a slightly different way - it looks for a kid from a diversity pool that is the most unlike its current set of acceptees that has also listed the school as one of their seven picks. Within that pool, it will pick the kid with the application where the school is ranked highest.

    It then recalculates its diversity index (which may change which type of kid is optimal for balancing its diversity index), and then repeats.

    ReplyDelete
  24. 10:22, if you are lucky enough in the lottery to be assigned to your #1 and #7 picks, then you will be assigned your #1. Really. That is how the alogorithm works. I suggest you write directly to Vicki @ PPS if you doubt this--she sat down last year and went over the program with one of the people who designed it.

    What you do should depend on how much you have your heart set on that #1 school. If you would really be fine with #7, then by all means list it. Then you can breathe a big sigh of relief in March when you have an assignment you like (and you can still go in the waitpool for your #1--if it is truly not a trophy school, it's possible you'll get the call). But if you really, really want that #1, and also have the fortitude to brave the waitpool process, then your strategy is better. (I personally would go with the former, but everyone has their own mileage.)

    Also, if you are English-only, preschool, not-poor, then you are probably not in the biggest pool for the district overall. A majority of our kids qualify for reduced/free lunch, after all. In terms of Clarendon applicants, yes, you would likely be in the majority. At a number of other schools, depending on where the computer is in the process as it churns through the algorithm, that is not necessarily the case. Marshall, Paul Revere, Daniel Webster, Monroe, Glen Park, Rosa Parks, Jose Ortega .... all interesting schools where your diversity profile might actually be an advantage at some point in the process.

    ReplyDelete
  25. 11:22, ... I think we are saying the same thing, that the computer will overlook a less desirable applicant (with regard to diversity) and choose someone who may not have put it highest but IS in a not yet adequately represented pool, right? And it will choose the kid who ranked it highest in his pool.

    That's why even with a school that has 33 spots and only 33 first choice applicants, they won't all get that school unless they happen to represent the right balance of the 16 pools. I just want to make sure I understand the process.

    Also, does the algorithm take into account the total number of applicants per pool with relation to each other and then attempt to reflect those percentages at each class, or does it go for a 50/50 split in each category? (I'm not even sure I understand what the50/50 split is, but I have read about it.)

    How does the system make sure that all pools are correctly represented based on each one's size compared to the other pools. For example, let's say there are 400 '0000' kids and 100 '0100' kids. Will the computer attempt to have 4 of the 0000 kids to every 1 of the 0100 kids in each class? Or will it go 1 for 1? Thanks for any clarity on this.

    ReplyDelete
  26. 10:22 AM here. Thanks for your response; it makes perfect sense and is very helpful.

    One more question, or a confirmation really : I assume that the diversity algorithm applies to the entire student body as a whole. In other words, it's all about adding diversity to the entire school, not by individual class year or anything else, correct ?

    ReplyDelete
  27. 10:22 AM again. I think I answered my own question from the SFUSD web site : apparently the diversity algorithm is per class year, not for the entire student body. Correct ?

    ReplyDelete
  28. 10:22am you are correct, it's per class, per year, not the whole school.

    ReplyDelete
  29. 12:05, re:
    "Also, does the algorithm take into account the total number of applicants. . . etc." No it's 50/50 on the binary diversity index.

    "How does the system make sure that all pools are correctly represented based on each one's size compared to the other pools." It doesn't. That's not how it is designed.

    ReplyDelete
  30. 11:54, re: "Also, if you are English-only, preschool, not-poor, then you are probably not in the biggest pool for the district overall. A majority of our kids qualify for reduced/free lunch, after all." You are right that majority kids do qualify for free lunch but unfortunately most don't turn their applications in on time so are not considered in Round 1. Even if they are siblings, they have to turn in the form and sadly most don't. PPS is doing more outreach on this.

    ReplyDelete
  31. 2:01, as someone else said earlier, just when you think you understand the algorithm, it vexes. Can you explain the 50/50 thing? I'm confused. If there are 16 pools, is it trying to have 50 percent of each in each school/class? My brain hurts.

    ReplyDelete
  32. 2:50: yes, they are trying for 50/50 in each category. They don't succeed of course, because they are limited by the makeup of each school's applicant pool, but that is how it is designed.

    I suppose you could argue that it should be set up to mirror the exact makeup of the district across these factor--or to mirror the makeup of children in the city, although that would be difficult as so many upper/middle class kids go private. Plus, that sounds even more extremely complicated! So they default to the binary. The overall district numbers for several of the most important factors, as I understand it, are not that far off 50/50 anyway.

    FWIW, I think Aptos Middle School may most closely mirror the makeup of the district, and has shown improved performance for all sub-groups, which is remarkable for a school with so many significant sub-groups.

    ReplyDelete
  33. Actually Aptos Middle School is very representative by ethnicity/race and socio-economic factors of SFUSD, but not at all of the ethnicity/race and socio-economic factors of San Francisco. What this means is if you are Caucasian (which is 49% of SF according to the census) and non-poor, no matter what school your child is in you will have your child in a more diverse environment than San Francisco as a whole, which is pretty great in terms of educating children in environments where people are not all like them!

    ReplyDelete
  34. I agree, 4:58, but I think it's worth pointing out that the caucasian or "other white" population of CHILDREN in SF is more like 24%, not 49%. You are right that the population as a whole, including adults and children, it is 49%. In SFUSD it is about 10%.

    My kid goes to Aptos by the way, and has plenty of friends of many backgrounds, and seems utterly unfazed by the great mix or by being in the minority. There's a cosmopolitan air about the place in that sense. It is cool to be part of the mix.

    ReplyDelete
  35. 11:58 - an other non-white would be a minority and enjoy a "cosmopolitian" mix at any school in San Francisco.

    ReplyDelete
  36. 6:43--at any public school, you mean.

    ReplyDelete
  37. "That's why even with a school that has 33 spots and only 33 first choice applicants, they won't all get that school unless they happen to represent the right balance of the 16 pools. I just want to make sure I understand the process."



    "How does the system make sure that all pools are correctly represented based on each one's size compared to the other pools. For example, let's say there are 400 '0000' kids and 100 '0100' kids. Will the computer attempt to have 4 of the 0000 kids to every 1 of the 0100 kids in each class? Or will it go 1 for 1? Thanks for any clarity on this."

    AFAIK, the four diversity indices are somehow combined into one (using some kind of vector algebra), and the target for that combined variable is 0.5. I'm uncertain whether the target for each diversity index is normalized against what the applicant population is (i.e. if 10% of the population are "1" for "in extreme poverty", that the target for that index is 0.1). I vaguely remember hearing they don't normalize the indices against the applicant population (which would be a big flaw in the algorithm), but an email to PPSSF would clear this up.

    ReplyDelete
  38. "But if some people end up with their #6 or #7, as people certainly have on this blog, how can it be true that it's #1 or 0/7?!"

    Depends on diversity criteria. I know one family that got AFY despite putting it #2, and they're still scratching their heads figuring out how that could have happened, given there were 4 first choice applications for every slot at AFY.

    "If that is true, can I just say that it would make a heck of a lot more sense to take all those who don't get their #1, and make their #2 into a #1 for a second pass, changing all the #s accordingly, and going on down like that?"

    Because that would raise issues of gaming the system, I suspect.

    There's a lot more strategy now than there was when it was thought that ranking made no difference to the probability of getting in. I think the effects will be benign (as more of the good but non-trophy schools will higher rankings), but it is more complicated.

    ReplyDelete
  39. "I suppose you could argue that it should be set up to mirror the exact makeup of the district across these factor--or to mirror the makeup of children in the city, although that would be difficult as so many upper/middle class kids go private. Plus, that sounds even more extremely complicated!"

    They could normalize each diversity variable against what the proportion in the overall applicant population for that grade is. Let's say 4,000 applications are received for kinder, and 1,200 receive some form of public assistance. Then the target for that variable should be 30%, rather than 50%.

    ReplyDelete
  40. @6:43, that is true for any (public) school, but it is also true that some schools are more diverse than others, with more representation from a variety of groups instead of one or two. This may or may not be important to people, but I think it is a nice thing to be a in classroom in which everyone is a minority and there are many minorities.

    I believe Harvey Milk is like this on the elementary level, and yes, Aptos is too for sure, also Hoover? Not so much James Lick and Giannini, Horace Mann or Everett or Roosevelt. Again--there are certainly other factors to consider, but the "cosmopolitan" factor where many sub-groups represent the huge diversity of this city (Chinese, Filipino, Vietnamese, Latino cultures, African American, Palestianian, South Asian, white kids of various cultural and national backgrounds) I've seen only at a smaller group of schools. Just something to think about.

    ReplyDelete
  41. "Depends on diversity criteria. I know one family that got AFY despite putting it #2, and they're still scratching their heads figuring out how that could have happened, given there were 4 first choice applications for every slot at AFY."

    Rank doesn't matter while they are able to balance. If #2 or #7 balances better than #1 (at a particular moment in time), then #2 or #7 is assigned.

    ReplyDelete
  42. I would say that Rooftop has a perfect diversity. No one group is a majority in my daugther's K class, and most "groups" are represented with a few kids. Maybe not the exact mirror of the composition of SFUSD, but very even across groups.

    ReplyDelete
  43. Who gives a crud about diversity in elementary school? Seriously. We live in an incredibly cosmopolitan city. This isn't Utah. People live here for a reason. It would take a whole lot of work to try to raise a sheltered kid here. It is so repulsive how much energy in SF is drained away from what matters: a caring and educational experience, and is wasted on this artificial notion of balancing racial groups..

    ReplyDelete
  44. "Who gives a crud about diversity in elementary school? Seriously. We live in an incredibly cosmopolitan city. This isn't Utah. People live here for a reason. It would take a whole lot of work to try to raise a sheltered kid here."

    You need to tour more schools to see how radically the socioeconomic and racial balance shifts from school-to-school. There's a world of difference amongst E.R. Taylor, Lilienthal, Moscone, Alice Fong Yu, and S.F. Community.

    "It is so repulsive how much energy in SF is drained away from what matters: a caring and educational experience, and is wasted on this artificial notion of balancing racial groups.."

    None of the diversity criteria are racial-based. Try to keep up.

    ReplyDelete
  45. Hi, nice post. I have been pondering this topic,so thanks for writing. I will probably be coming back to your site. Keep up great writing.

    ReplyDelete
  46. Hi!

    I'm a reporter with the Examiner and was hoping to talk to someone about the first round of applications.

    Anyone available to talk today?

    Please e-mail me at kkelkar@sfexaminer.com

    ReplyDelete
  47. can someone please list the infamous and untouchable eleven schools? we're deciding on our list tonight.

    Thanks

    ReplyDelete