A recent report from Mackinac Center for Public Policy seeks to measure and publicize high school performance by ranking schools according to their test scores after attempting controlling for students’ “economic status.”
Associate Professor John T. Yun of Michigan State University reviewed The Michigan Context and Performance Report Card: High Schools 2018. He concluded that while the stated goal of the report is laudable, the reality falls far short due to several shortcomings.
The report touts its
decision to take free-lunch status into account as its major
contribution, in comparison to past school rankings (although this type
of calculation has been done previously in many contexts). Beyond this
choice, however, the study lacks both justification for and explanation
of its methodological decisions. The validity and reliability of
combining disparate tests across different years without proper equating
invalidates the findings – particularly for the high-stakes
applications presented in the report.
Additionally, the
free-lunch percentage measures used in the study have a great deal of
measurement error, which argues against this sort of ordinal ranking.
And the use of a single predictor with unacceptably low correlations for
this type of usage grossly oversimplifies and biases the estimates.
Given these shortcomings,
the rankings presented in this report should be given no weight in any
discussions of policy or practice. In fact, this report does a
disservice by introducing questionable information in an easily readable
form that is not substantiated by any credible analysis.
Find the review, by John T. Yun, at:
Find The Michigan Context and Performance Report Card: High Schools 2018, written by Ben DeGrow and Ronald Klinger and published by Mackinac Center for Public Policy, at:
No comments:
Post a Comment