The rating scale that defines how the evaluator should score agent performance, either numerical or categorical.