Small-school students do better in NYC

Students at New York City’s small high schools earn higher test scores and are more likely to graduate in four years, according to a MDRC study.  The analysis compared students who won a lottery for admission to a small high school with others who applied but lost the lottery.

The latest findings show that 67.9 percent of the students who entered small high schools in 2005 and 2006 graduated four years later, compared with 59.3 percent of the students who were not admitted and instead went to larger schools.

. . . This increase was almost entirely accounted for by a rise in Regents diplomas, which are considered more rigorous than a local diploma; 41.5 percent of the students at small schools received one, compared with 34.9 percent of students at other schools.

Small-school students earned higher scores on the English Regents exam, but there was no difference on the math Regents.

Here’s a link to the policy brief pdf).

About Joanne


  1. Small schools, bought and paid for by Bill Gates and a study he paid for seeming proves his ideas work. Hmm, doesn’t sound suspicious at all.

    As always, the “study” is not published in a peer-reviewed journal where findings are checked for accuracy and legitimacy.

    • Roger Sweeny says:


      Peer review does not check findings for accuracy and legitimacy. Before an article is published in a peer reviewed journal, it is sent to (usually) two people who the editors consider expert in that area and who have agreed to read it. The experts look to see if anything is obviously wrong and if they think the proposed article will add something to the sum of human knowledge. Peer reviewers are generally not paid and spend less than an hour on the process. There is almost never any checking of the assertions in the article or rerunning of any of the statistical tests.

      Lots of bad articles make it through the peer review process, and some good ones don’t.

      BTW, since most government studies are published by the agencies themselves, almost none of them go through the peer review process. Government employees, like everyone, have agendas. Do you dismiss all government studies?

      • Roger,

        This “study” was not a govt. publication, but a study bought and paid for by Bill Gates to verify that his small school idea, which his foundation has already abandoned as a failure, is successful.

        Despite the author’s claim it was a rigorous study I was unable to find a link to the study itself, just a policy brief and an overview.

        If its so rigorous then why not let people look at the entire study, or submit it for publication instead of self-publication with a lack of details?

        • Roger Sweeny says:


          I did not mean to imply that this was a government study.

          As far as I’m concerned, a “policy brief and overview” that does not contain a link to the study is not worth taking seriously. People should be able to look at what the study author did and draw their own conclusions.

          I’d actually go even further. As a condition of publication, journals should require authors to put on the web any data that they used in a study and any non-obvious manipulations they did of it.

          • Roger,

            Then we are in agreement. However the “working paper” is a favorite tactic of the “reformers”. It is released to the public as gospel, with people like Joanne and most journalists too lazy to do any follow up on it, and printed as the truth.

  2. Did they properly control for the difference in variation between smaller and larger schools this time? Or in other words, check to see if small schools are also disproportionally represented at the bottom of the score distribution.