This paper presents the latest in a series of studies examining school district report cards for their usefulness planning and implementing school improvement. The 1992-93 Arkansas report cards were studied. These were similar to the report cards from Tennessee that were the subject of the initial study in that they used and reported a norm-referenced national test and a criterion-referenced state-designed achievement test. Demographic items were used as independent variables, while student outcome items were used as dependent variables. These six outcomes were connected with test results on the Stanford Achievement Tests, the American College Testing program test, and the Arkansas Minimum Performance Test. Findings demonstrated that basic statistical techniques can yield misleading ideas about relationships among school and community characteristics and student achievement. More sophisticated statistical treatments are needed to portray relationships more accurately. Different school and community characteristics had dramatically different effects on the six outcome indicators reported. Arkansas report cards needed more information in terms of outcome indicators and categories for weaker student performance. In Arkansas, it is apparent that the impact of black students on overall student outcome data needs more study and clarification. Nine appendixes present detailed tables of the statistical relationships among variables. (Contains 11 tables and 13 references.) (SLD)