Tag Archives: KIPP

Welcome to the Profit of Education website. Continuing the conversation begun in the book Profit of Education, we discuss the latest economic evidence on education reform.

More on KIPP schools and a little on methodology

A new article in Education Finance and Policy, Do KIPP Schools Boost Student Achievement?,¬†offers a nice opportunity to say something about methodology in program evaluation. First though, a bit about the substantive results. KIPP schools are probably the best known … Continue reading

Share
Posted in Uncategorized | Tagged | 6 Comments

6 Responses to More on KIPP schools and a little on methodology

  1. sjl says:

    This conclusion is so wrong. If there is important selection into KIPP on unobservables that affect student outcomes (such as parental motivation), then a propensity score matching analysis demonstrates precisely nothing.

    • Dick Startz says:

      Suppose we modestly expand the KIPP program. Presumably students will select into the new schools in the same way they select into the current schools. The evidence is that these students will do better than if they did not have KIPP schools to attend.

      • sjl says:

        It is not. The marginal kids in the KIPP expansion will be those who had been at the top of the non-KIPP distribution previously. You just don’t have an appropriate counter-factual here.

        • Dick Startz says:

          Very few students have access to KIPP schools. There could be considerably increased KIPP access without changing the margin.

          • sjl says:

            I can’t see what relevance this has. The point is that, if there is selection on unobservables, the KIPP Students would have performed better than the average student with their observable characteristics in a non-KIPP environment as well–that’s what the term means. This implies that, without some analysis of the size of the resulting bias, you can’t rule out the possibility that the return to KIPP is zero for all students. You may believe your conclusion to be true, but it doesn’t follow from the study you are telling us about.

Leave a Reply

Your email address will not be published. Required fields are marked *

Who Benefits from KIPP?

Evaluating the performance of a school can be a tough proposition because motivated parents may be more likely to send their kids to a school with a good reputation, so results reflect getting easy-to-teach kids rather than what the school … Continue reading

Share
Posted in Uncategorized | Tagged , , , | 1 Comment

One Response to Who Benefits from KIPP?

  1. Bob Roseth says:

    It’s exactly for this reason that a carefully-crafted charter school law, such as the one that is on the ballot in Washington state, makes sense. While the Stanford/CREDO study can be read as saying that, on average, charters are no better than garden-variety public schools. But if you look more deeply, as this study suggests, some charter school organizations have developed models that, on average, produce results significantly above average — and that this bodes well when their target population is those at greatest risk. This is just one strategy for closing the achievement gap, but one worth having.

Leave a Reply

Your email address will not be published. Required fields are marked *

Charter schools: KIPP in Lynn, MA

Do charter schools work? Getting good statistical evidence is tough because parents who work to get their kids into a charter school are special (If you haven’t already, go see Waiting for Superman). We might expect their kids to outperform … Continue reading

Share
Posted in Uncategorized | Tagged , , , , , | 1 Comment

One Response to Charter schools: KIPP in Lynn, MA

  1. MS says:

    How much can we legitimately extrapolate from the effect of KIPP to the effect of charter schools in general? Is there any good research out there on the variation in charter school quality?

Leave a Reply

Your email address will not be published. Required fields are marked *