GothamSchools — daily independent reporting on NYC public schools

guest perspective

The Hoxby study: Big findings, big omissions

The charter school study released this week that suggests charter school students are catching up with their suburban peers leaves many questions unanswered. As GothamSchools and the New York Times reported, the study found that a student who attended a charter school for all of grades K-8 would close approximately 86 percent of the “Scarsdale-Harlem achievement gap” in math and 66 percent of the achievement gap in English.

But here’s what I want to know.

1. How many students in the study actually completed grades K-8 in charter schools? Nowhere in the study does it say. Yet these students supposedly close 86 percent of the Scarsdale-Harlem achievement gap in math and 66 percent in ELA. It would help to know how many there are.

2. How many students were in charter schools for two, three, four, five, or more consecutive years? Nowhere in the study does it say.

3. For that matter, how many students in the study were in charter schools? Nowhere in the study does it say. It states that “the current report covers 93 percent of New York City’s charter school students who were in test-taking grades between 2000-01 and 2007-08,” but it does not give numbers for those grades.

4. How many charter school students took tests each year? Nowhere in the study does it say. It gives the number of charter school applicants who took tests in each grade between 2000-01 and 2007-08 (combined), but it does not break down these numbers into “lotteried-in” and “lotteried-out” categories. Nor does it say how many took the tests each year, or how many took them in charter schools for consecutive years.

5. How many students (charter or public school) in the study switched schools within the same school type, e.g., charter to charter or public to public? Nowhere in the study does it say. This is important. We should consider the difference between stable and unstable learning conditions.

6. How many students who started out in the study were no longer in it by the end? About 14 percent of charter school students in the study transferred to regular public schools. It appears that another 24.8 percent of students left the study for other reasons (moving to the suburbs or transferring to a private school; graduated from high school; etc.). It is not entirely clear whether these two groups are discrete. If they are, then nearly 40 percent of charter school students in the study did not remain in charter schools for the duration of the study.

7. How many charter schools have existed long enough for a student to have attended grades K-8 (or K-5, or as long as the school extends)? Nowhere in the study does it say.

8. Which schools did the “lotteried-out” students attend? Nowhere in the study does it say. If they were in troubled, overcrowded, or dangerous schools, then this could affect the results.

We are left with many unknowns. We do not know how many students remained in charter schools for two, three, four, or more consecutive years. We know nothing about the schools that the “lotteried-out” students attended; we don’t even know whether they remained in the same public schools for the duration of the study. But if I were to raise my hand and ask one (two-part) question, I’d ask (a) how many students were in charter schools for grades K-8 and (b) their ratio to the total charter school population. If they are a hypothetical group, the study should make that clear.

  • http://www.specialeducationmuckraker.com Dee Alpert

    At some point in the far, far future, we will have audited, verified numbers from NYC DOE schools re real enrollment, actual attendance (not ATS default “present” attendance), legitimate transfers to non-NYC DOE schools, and real moved-out-of-NYC kids, as well as real numbers of dropouts including those who cease attending school during or immediately after middle school. It is my understanding that the latter number is substantial.

    The NYS Ed. Dept. has a unique student identifier system which has been functional for quite a few years now. This could be used, if the NYC DOE wished to do so, to provide some serious verification for reported numbers on each of the above issues and would also tell us how many kids should have, but were not, tested at each appropriate age and grade level. The fact that this information is available, but has not been made public, leads me to believe that were it to be made public in an unmanipulated form, there would be a stupendous scandal.

    Be that as it may, until we can get this data for the NYC DOE and its individual schools – ALL of them, including District 75 special ed. schools – then it is not particularly appropriate to ask for this information only for charter schools.

    It would also be useful if the NYC DOE released the hard numbers on how many kids in each school, in each grade, were given last minute modifications and/or accommodations due to some alleged temporary infirmity as well as the same for kids marked “absent” or not tested due to some clerical error. Again, my understanding is that these are not-insignificant groups. Do charter principals find instant infirmities more frequently than their regular public school principal-peers? Interesting question, albeit one to which our grandchildren may be able, finally, to get some answers.

  • http://ljohnson562@charter.net Linda/Retired Teacher

    How were these students evaluated? Did the Stanford researcher do her own testing or did she accept scores from the schools attended by the students?

  • Michael M.

    L/RT,
    I hear ya, but on this I am inclined to cut slack.
    Just imagine the hue and cry if, rather than piggbacking off the same test-and-score data used by other researchers, Hoxby had ginned up her own.

  • http://ljohnson562@charter.net Linda/Retired Teacher

    Mike, yes I do see what you mean, but these researchers are basing all their work on test scores that are likely to be invalid. I’m surprised universities don’t insist on more accurate data.

  • Caroline

    Sorry to be redundant, everyone, because I’ve posted this on other threads. But just for the benefit of anyone reading only this thread:

    Caroline Hoxby, who headed this study, is not an impartial reseacher, and it’s misleading to represent her “study” as credible academic research. She is well-known as a longtime advocate of “free-market solutions” in education — charter schools, privatization and vouchers. It’s a glaring lapse on the part of all press who reported on this so-called “study” that they did not identify it as having been conducted by an openly partial advocate of charter schools.

  • http://ljohnson562@charter.net Linda/Retired Teacher

    It’s so difficult to control critical variables in these studies. For example, a friend of mine who teaches at a charter school told me that they basically drill the children on the actual test from September to May. Without oversight of these schools, how do we know that the charter schools aren’t just doing a better job than other schools at “teaching to the test?” Yes, I know the regular public schools are doing the same thing but perhaps not to the same extent. That’s the problem. We really don’t know what these research studies mean. Frankly I am surprised that institutions like Stanford and Harvard don’t insist on more reliable data for their studies. Basically they should only agree to do studies in cities that can guarantee a reasonable amount of security for their standardized tests.

    (Yes, I know I’m probably being naive.)

    I am not really arguing with Professor Huxby’s results because I am almost certain that charter schools attract a select population; therefore I’d expect the test scores to be higher just as they are in almost any school (parochial, private) where parents select. This reflects the importance of parental involvement, quite possibly the most critical variable of all. Also, I know from my experience in low-performing CA schools that at least a third of the children are at or above grade level. Their parents are the ones who would apply to a charter school. It’s obviously important to control these studies carefully because researchers could easily come to the wrong conclusions. In the case of this study, Prof. Hoxby has come to the conclusion that charter students do “better” (higher test scores) than students who applied and didn’t get in, but could it be that these charter schools drilled more on the test because of a lack of oversight? There are many other unchecked variables as listed by Diana.

    An educational infrastructure is being built on the strength of test scores that are very “iffy,” to say the least. I hope the universities band together to insist on more accurate data. By doing so, they can really provide an important service to the children of our country. If we continue with this smoke and mirror approach to “reform,” we might find ourselves with another failed, but expensive, educational fad with nothing to show for it.

  • Greg

    While I agree that Hoxby should release the N’s for the study so we know exactly how big the sample size is, it is still a silly list of questions or line of attack because an expanded data set will likely further enhance the positive results for charters.  The number of excellent charters that have opened since 2005 is far larger than those in 2001-4 (all the top charters in fact, Williamsburg, AF, Infinity, Democracy Prep),  and the idea that because the students were lotteried into bad schools is exactly the point of the study.  

    The issue of mobility and persistence rates is also a red herring.  if you look at mobility in traditional schools in districts like Harlem and Bed Stuy, it is off the charts…20%+ A YEAR.  so if charters have a cumulative 40% rate over 7 years, they’re way below the district mobility.  But then again, charter critics would say that proves “Creaming.”  In fact, that proves that a family who has a good school is LESS likely to move because they are satisfied than one who goes to a school where their child is unsafe. 

  • Michael M.

    Greg,
    Some critics of the study have specifically pointed to what you are dismissive of in para one: Comparison of charter schools to bad schools. Got “mission creep?”

    Next, proof of “creaming” doesn’t depend on mobility and persistence rates.

  • http://ljohnson562@charter.net Linda/Retired Teacher

    Here is a specific situation that might be part of the Hoxby study: Mrs. Jones wants to get her son out of his terrible urban classroom because of low achievement and many disruptions. Mrs. Smith feels the same way. Although they are poor, they are both very concerned about the education their children are receiving. Both of them apply to the local charter where their children are placed in a lottery. Mrs. Jones’ son gets in but Mrs. Smith’s child is placed on the waiting list for over a year. During this time Mrs. Smith is forced to send her child back to the school with all the disruptions. Whose child will have a more successful school year? Do we really need a Stanford study to give us the answer to this question? Would the fact that all the parents have chosen the charter school affect the outcome for their children? Do you think Mrs. Smith’s child would do better in the traditional public school than the children’s whose parents are less involved? Actually these questions were all answered over 40 years ago. Perhaps it’s time to recognize the importance of the parents and apply that knowledge.

  • Diana Senechal

    The study emphasizes the partial closing of the “Scarsdale-Harlem” gap. But the students who have made these gains may be a small portion of the charter school population. In any case, they are a subset with particular characteristics. And they are being compared to “lotteried-out” students as a whole, not a subset of them.

  • http://ljohnson562@charter.net Linda/Retired Teacher

    I just received my copy of Education Week today and discovered that Professor Hoxby used children who were randomly assigned to the charter schools instead of children whose parents applied. This makes a huge difference in the validity of the study and does seem to indicate that perhaps the students in the charter schools performed much better. I’m still very much a cynic so I would want to know if the charter schools drilled the students on the test more than the regular school teachers did. After all, without public oversight, who would know?

    I must admit though, that there is a strong possibility that the charters are improving achievement for low-income minority students. If this is indeed the case I hope teachers apply to open and manage their own charters. This would give them the opportunity to have autonomy over the important work that they do.

  • Caroline

    And the study is still discredited because it was conducted by a vocal longtime advocate for charter schools, without disclosure of that fact. Sorry to repeat myself, but I’m doing my best to make sure this breach of standard, ethics and honesty is not forgotten.

    If the charter advocates have to be dishonest to boost their cause, shouldn’t they really be questioning whether their cause is as worthy as they want to think?

  • http://ljohnson562@charter.net Linda/Retired Teacher

    Caroline,

    I understand what you are saying because when I was in graduate school I saw how researchers “guided” their studies to make results come out a certain way. With endless numbers of variables involved, it’s not at all difficult to do. However, I find it hard to believe that a Stanford researcher would spend eight years on a study if she weren’t trying to get valid results. Still, she definitely should have told journalists about her charter school bias. It’s all very puzzling to me. Since charter schools seem here to stay, do you agree that teachers need to take the lead? What bothers me more than anything is the fact that non-teachers are opening these schools and establishing leadership that ought to go to teachers.

  • Caroline

    If you Google Hoxby, there’s a history of disputes over her past studies. Maybe that’s par for the course for any researcher in a controversial area, but her stature as a vocal, partisan, longtime advocate of charter schools (and vouchers, in relation to previous studies) certainly comes into it.

    It was dishonest of her — and Stanford — not to disclose her background, and it was unforgivably shoddy and lazy of every reporter/editorial writer/etc. who wrote about this not to learn about her background and include the disclaimer. Failing to do so misleads the reader, which is the worst sin a journalist can commit.

    All that said, it may be a perfect, pure-as-the-driven snow study, but the disclaimer would still be called for by any ethical standard.

    Given the amount of dishonesty that props up the charter hype, I’m not that convinced that they’re here to stay. If I didn’t see so much deliberate dishonesty, I might believe otherwise. Perhaps the dishonesty will eventually eat away at the aspects of the charter movement that engage in and benefit from it, leaving some cleaner version. I can’t picture it. But any movement that has to engage in so much dishonesty to boost itself is not destined for long-term stability or success, in my view.

  • Michael M.

    L/RT,

    The parents applied. THEN there was an albeit random lottery.

  • Caroline

    Well, there’s a charter high school here in San Francisco that requires an enrollment application much like that of a private school, including teacher recommendations, an essay, transcripts, signed parent commitments. Then they insist that their drawing of the ones they put into the hat (or whatever it is) is purely random. Of course they don’t even try to claim that they put all the applications into the receptacle for the drawing, not to mention the tiny issue that the required documents might have a little teeny creaming effect.

    The school is Gateway High School, San Francisco, and anyone can check out their application process on their website.

  • http://ljohnson562@charter.net Linda/Retired Teacher

    This is what was written in Education Week:

    “Studies comparing student achievement in charter schools with that in regular public schools are difficult to do credibly, however, because applicants and their families are presumably more academically motivated than students who remain in regular schools. Ms. Hoxby was able to overcome that problem in New York City because its charter schools are oversubscribed, requiring administrators to us lotteries to allocate seats. That allowed researchers to compare students randomly assigned to charters with students who applied but did not win a seat in a charter school, thus providing for a more apples-to-apples comparison.”

    If Hoxby took students whose parents applied for a position, how would that alter the effect of “the more academically motivated?” I don’t get it. One journalist who seems interested in getting at the truth is Bill Turque of the Washington Post. Maybe he’ll do some digging.

    Oh, my lights just went on. Hoxby figured that her control group also were among the “academically motivated.” That doesn’t prove anything important to me because if you take a group of motivated kids and place them in a school together, that school is going to be better. That’s the secret of the Catholic schools. The children who were not placed were presumably spread out to different schools so that would certainly affect their education significantly. To truly control that group, you’d have to put them all in one regular public school that had no other students. Also, what we all want to know is this: Can a charter school take the SAME kids that are in a regular public school and get superior results? If they can prove that, I will become a supporter.

  • Diana Senechal

    The wording of the Edweek article is slightly misleading. The students were “randomly assigned” to charters insofar as applicants were selected by lottery. That is, they applied. The real problem of this study is that it pertains to a subset (perhaps a small one) of the actual NYC charter student population. It is the students who have been in charters for grades K-8 that supposedly have gone far in closing the “Scarsdale-Harlem achievement gap.” But nowhere does Hoxby state the ratio of these students to the entire charter school population. Nor does she say whether these students are concentrated in certain schools. We need more information, period.

  • canwetalk

    Diana,
    Thank you for your analysis of the Hoxby study (I rather call it the Hoaxby study) and your thought-provoking questions that should raise concerns about the study’s lack of information and true accuracies.

  • http://ljohnson562@charter.net Linda/Retired Teacher

    Ever since the “Texas Miracle” we have been in a period of lies, fraudulent testing, “studies” based on the fraudulent testing, and journalists who aren’t asking questions. Let’s hope it all changes soon.

Tips, questions, feedback?

Contact us at .

Follow GothamSchools

RSS

Feb. 10: You’re invited!

Chalk It Up

Recent Comments

45 comments so far today

Our Twitter Updates

  • ” you don't want to come to class? Have a packet. You don't like your teacher? Have a packet” - @leoniehaimson 1 hr ago
  • .@leonileoniehaimson brings letters from anonymous teachers with damning tales.of credit recovery: giving out CR ”packets” like skittles.. 1 hr ago
  • At credit recovery town hall hosted by Regents. Testimony so far by principal, and 2 former teachers. Principal support; teachers critical 1 hr ago
  • Our report about the city's decision to keep two schools open, complete w/ co-location worries & political speculation: http://t.co/RO59PMh1 2 hrs ago
  • Chancellor Walcott about Wadleigh and KAPPA VII: "While these two schools continue to struggle, what we learned is that they are also... 3 hrs ago
  • More updates...

Archives

February 2012
M T W T F S S
« Jan  
 12345
6789101112
13141516171819
20212223242526
272829  
?>