×
I know that the Journal Sentinel decided to ignore the Legislative Audit Bureau's review of the latest study on voucher schools (the paper is too busy propping up the rumored campaigns of Paul Ryan and Tommy Thompson, apparently). But the study is newsworthy anyway.
To recap: when folks started questioning the performance of voucher students a few years ago, the voucher supporters decided that if they picked some researchers to study the program over five years then they'd get some positive results. So the “privately funded” (ahem) School Choice Demonstration Project, based at the University of Arkansas, was born.
The Arkansas researchers demographically matched kids who attended voucher schools and MPS and then attempted to follow them through the years. The catch, of course, is that the researchers don't release data on individual schools, so policy makers and parents can't figure out which schools are better than others. (Last year, however, Democrats began requiring voucher schools to take the Wisconsin Knowledge and Concepts Exam and then have DPI release the results for individual schools.) The longitudinal study is intended to end after the 2010-2011 school year results are in.
But the researchers hit a few snags along the way.
First, the researchers continually found that there was no meaningful difference between the test scores of voucher and MPS students. So the voucher supporters weren't too happy about that (privately, at least—in public they spun the results in their favor).
The latest LAB report, which reviews the researchers' data, found the same result:
But the new report adds another wrinkle to the study's problems. The LAB confirmed what we already knew: that there's a high rate of transfers from voucher schools to MPS and elsewhere. The problem is that the rate of transfers out of voucher schools is so high that it threatens the study itself, since there aren't enough kids to match up.
Only 41.3% of the 2,727 choice pupils in the researchers' original sample remained in the choice schools in the 2009-2010 school year.
That means that almost 60% of choice students have bailed within the past four years. Wow. In contrast, 71.6% of MPS students in the sample stayed put in MPS during the past four years.
The LAB found that 32.8% of students in the voucher sample transferred to MPS, while only 6.8% of MPS students transferred to a voucher school during the same period.
The LAB wrote that the high rate of transfers makes it difficult for researchers to determine whether a students' performance can be chalked up to what he or she learned at a voucher school, or whether it's due to what the student learned at MPS.
The LAB wrote that its results varied a bit from the original researchers' results because the Arkansas team “used statistical techniques to attempt to compensate for missing test score data.” The Arkansas researchers failed to note that and didn't provide the LAB with more detailed info on how the test scores were estimated.
So it looks like the performance of voucher schools isn't the “Wisconsin miracle” that its supporters had hoped for. And given the choice between vouchers and MPS, parents and kids prefer MPS. Unfortunately, the Walker administration has decided to expand this program to private schools in the rest of Milwaukee County and Racine. Parents, you've been warned.
To recap: when folks started questioning the performance of voucher students a few years ago, the voucher supporters decided that if they picked some researchers to study the program over five years then they'd get some positive results. So the “privately funded” (ahem) School Choice Demonstration Project, based at the University of Arkansas, was born.
The Arkansas researchers demographically matched kids who attended voucher schools and MPS and then attempted to follow them through the years. The catch, of course, is that the researchers don't release data on individual schools, so policy makers and parents can't figure out which schools are better than others. (Last year, however, Democrats began requiring voucher schools to take the Wisconsin Knowledge and Concepts Exam and then have DPI release the results for individual schools.) The longitudinal study is intended to end after the 2010-2011 school year results are in.
But the researchers hit a few snags along the way.
First, the researchers continually found that there was no meaningful difference between the test scores of voucher and MPS students. So the voucher supporters weren't too happy about that (privately, at least—in public they spun the results in their favor).
The latest LAB report, which reviews the researchers' data, found the same result:
“The project's five-year longitudinal study shows no significant difference in the performance of choice and similar MPS pupils after four years of participation, and only one year remains to determine whether participation in the choice program affects academic achievement.”
But the new report adds another wrinkle to the study's problems. The LAB confirmed what we already knew: that there's a high rate of transfers from voucher schools to MPS and elsewhere. The problem is that the rate of transfers out of voucher schools is so high that it threatens the study itself, since there aren't enough kids to match up.
Only 41.3% of the 2,727 choice pupils in the researchers' original sample remained in the choice schools in the 2009-2010 school year.
That means that almost 60% of choice students have bailed within the past four years. Wow. In contrast, 71.6% of MPS students in the sample stayed put in MPS during the past four years.
The LAB found that 32.8% of students in the voucher sample transferred to MPS, while only 6.8% of MPS students transferred to a voucher school during the same period.
The LAB wrote that the high rate of transfers makes it difficult for researchers to determine whether a students' performance can be chalked up to what he or she learned at a voucher school, or whether it's due to what the student learned at MPS.
The LAB wrote that its results varied a bit from the original researchers' results because the Arkansas team “used statistical techniques to attempt to compensate for missing test score data.” The Arkansas researchers failed to note that and didn't provide the LAB with more detailed info on how the test scores were estimated.
So it looks like the performance of voucher schools isn't the “Wisconsin miracle” that its supporters had hoped for. And given the choice between vouchers and MPS, parents and kids prefer MPS. Unfortunately, the Walker administration has decided to expand this program to private schools in the rest of Milwaukee County and Racine. Parents, you've been warned.