Focuses on the applications of generalizability theory to performance assessment. This book examines the dependability of the scoring of large-scale performance assessments by humans. It presents principles in identifying common missing data patterns inherent in assessment designs.