The reliability of the pass/fail decision for assessments comprised of multiple components

2015 | Publications

The most serious consequences for a student in an exam are the decision of “pass” or “fail”. Like the measurement reliability of the scores, the reliability of the decision “passed” or “failed” needs to be determined in high-quality assessments.

Often, exams of one subject (credits) are made up of several sub-exams that must be passed independently. In this case of a “conjunctive” linking of the individual decisions “passed / failed” as well as in other complex existing rules, adequate evaluation procedures are necessary to estimate the accuracy and consistency of the decision “passed / failed”. To date, there are only a few papers available on this issue and a general methodology was published by Douglas and Mislevy in 2010.

The study aims to analyze the reliability of the decision “passed / failed” using a composite test in which several sub-exams have to be performed independently and discusses the implications for an improved methodology for identifying students who do not meet the minimum requirements.

Möltner A, Tımbıl S, Jünger J: Die Zuverlässigkeit der Entscheidung „bestanden/durchgefallen“ bei zusammengesetzten Prüfungen. GMS Z Med Ausbild 2015; 32(4):Doc42 (2015)