This report reviews the accuracy and reasonableness of statements in the print media about student proficiency on the 1990 National Assessment of Educational Progress (NAEP) in mathematics. It explores the apparent effects of two reporting approaches, the anchor-point method used by the Educational Testing Service and the National Center for Education Statistics since 1984 and the achievement-level (performance standards) method instituted in 1990 by the National Assessment Governing Board. In presenting anchor points and achievement levels the majority of writers incorrectly portrayed performance as discontinuous and ignored the continuum of success and failure. P-values, when provided, were frequently misinterpreted. Only a minority of articles mentioned the judgmental nature of the levels. Use of these metrics appeared to help the press by providing quotable and seemingly clear expressions of test results, but many articles were simplistic or incorrect, and important information often went unreported. Neither method as implemented for the 1990 results was adequate. Some suggestions are made for improved reporting. Two tables present details about the methods. (Contains 40 references.) (SLD)