It is argued that achievement test means, based on group scores, or other district-wide summary scores do not provide valid measures for diagnosing curriculum weaknesses nor for initiating curriculum changes. Group scores are probably valid for some purposes, but should not be used as criterion variables for evaluating the effectiveness of instruction or for indicating the quality of education a school district provides. Based on 11 case studies of district science instruction in kindergarten through grade 12, it was found that a lack of communication existed between the district curriculum coordinators and testing personnel, that curriculum coordinators paid little attention to test results, and that testing people were not concerned about curriculum review. In another case study, researchers found no instances of curriculum change or recommendations for change based directly on student testing. Using group test means as a basis for curriculum diagnosis or program evaluation is a procedure which is not grounded in validity studies, but is based on experience and intuition. (MH)