Compared student ability estimates with their true abilities under different rater scoring designs and differentially weighted composite scores through simulation. Results indicate that the spiraled rater scoring design without modeling rater effects works as well as the nested design in which rater tendencies are modeled. (SLD)