The need for improved measures is particularly acute in reading because, in spite of the magnitude of time and effort which continues to be invested in reading, there is no insurance that the outcome is indeed proportionate to the effort involved. How much of the educational system's performance relative to its own goals is measured by a standardized test is unknown. Yet this is the kind of information which must be available to a school system if it is to make sound decisions on the effectiveness of its programs. The longitudinal evaluation study using criterion-referenced measures of important reading-related skills which is briefly described in the report is seen as offering a new model for tests development which allows for some user involvement in the construction process, as well as contributing significantly to the solution of problems raised in the context of this report. (Author/RC)