Already in 2016, Judith Zimmermann defended her PhD dissertation at ETH Zürich on “Information Processing for Effective and Stable Admission”. Although the statistical handling of her data will no doubt have been flawless, her theoretical and conceptual framework leaves room for improvement – and criticism. She seems unaware of the crucial distinction between a) what students have to be good at, b) how good they have at it, and c) what tools can be used to ascertain this when deciding if a student will be admitted or rejected when applying for a Master’s programme.
She shows no awareness of the extensive body of expertise in the American graduate admission community and seems to not care about the issue of false negatives: rejected students who would have done well. This is remarkable, as recent findings at well-reputed American Graduate Schools show that this is indeed a problem. Through their National Student Clearing House system, American institutions can – and do – find out that their rejected applicants too often are admitted and successful at other schools that are as good – or better.