Journal article

A robust grade adjustment procedure

Assurance of equity in student outcomes includes ensuring that module results are comparable within reasonable bounds. We present a procedure that first compares (and adjusts if necessary) module variances. Then, student results are individually normalised to standard variates. A simple sign test is used to identify modules with disproportionately good or poor results. Module results are offset so as to adjust averages that do deviate. A detailed examination of an artificial data set shows that the proposed very simple procedure yields results that agree with a sophisticated statistical analysis.

Related material