Rater effects: the use of multilevel modelling to monitor raters

Funder: Pearson UK (£29,228); Principal Investigator: Professor Jo-Anne Baird; Collaborators: Dr Daniel Caro (OUCEA), Malcolm Hayes (Pearson), Alex McKee (Pearson); Kath Thomas (Pearson), Ed Wolfe (Pearson), George Leckie (Bristol).

The Pearson rater effects project investigated the value of multilevel modelling (MLM) in evaluating leniency/severity effects on the marking of students’ work and providing feedback to marking monitoring systems. Examination data from seven question papers provided by Edexcel were used for this purpose.

This is part of a larger project, the Marking Quality Assurance Project, in which the applicability of different statistical techniques to study rater effects was investigated.