Comparison of Inter-Rater Reliability Techniques in Performance-Based Assessment

The aim of this study is to analyse the importance number raters and compare results obtained by techniques based on Classical Test Theory (CTT) Generalizability (G) Theory. Kappa Krippendorff alpha CTT were used determine inter-rater reliability. In descriptive research data consists twenty individual investigation performance reports prepared learners International Baccalaureate Diploma Programme (IBDP) also five who rated these reports. Raters an analytical rubric developed Organization (IBO) as a scoring tool. show that statistical failed provide information about sources errors causing incompatibility in criteria. studies G provided comprehensive increasing would increase reliability values. However, raised idea it important develop descriptors criteria rubric.