The logic, i.e. the purposes, assumptions and implications of using performance objectives in evaluating schools is reviewed. The distinctions are recognized between standards for individuals and standards for groups as well as the distinction between normative and criterion-referenced standards. It is not appropriate to set standards for types of learning where cognitive development is a principal contributor to the type of capability being measured, if the intent of those standards is to try to get all students to achieve at a specified level. The integrated skills of reading and mathematics problem solving are examples. It is appropriate to set standards for acquiring finite bodies of information, or skills where their acquisition at the level designated is judged to be necessary for a well-defined purpose such as job competence, personal competence, social competence, or learning prerequisite to higher levels of education or training. (Author/RC)