The Role of Expectations in Set Size Evaluations