Interrater reliability and agreement of performance ratings: A methodological comparison