Approaches to data analysis used by the National Assessment of Educational Progress (NAEP) and possible alternative methodologies are reviewed. Since various components of the NAEP model affect analysis, key elements of that model, including procedures for developing objectives and exercises, drawing the sample, collecting data and reporting results, are briefly described. Performance on sets of items is summarized by using the average of percentages of acceptable responses to the items in a set. Methods for computing the statistics commonly used by NAEP and the standard errors of those statistics are described. The paper presents answers to questions frequently raised regarding NAEP analytical procedures, including such topics as the use of NAEP data to establish causal relationships, use of data adjustments, possibility of increasing the number of background questions, potential for setting performance standards, provision of different data breakdowns and new methods of analysis. For the 1979-80 reading assessment, data will be reported by achievement class, a new reporting variable. Appendices define reporting variables, present forms used to collect background information, describe population proportion smoothing methods, provide information about the problem of nonresponse in assessment samples, and describe methods of accounting for sample design with design effects. (Author/GK) Primary type of information provided by report: Procedures (Analysis) (Data Processing).