Comparison of student test scores between states, school districts, and even schools continues to be a popular measure of student achievement. However, these comparisons reveal little about the quality or effectiveness of educational programs, only the varying difficulty of educating different populations of students. This report uses U.S. Census data and information on Pennsylvania school districts to explain differences in the difficulty of the educational task. Demographic data on adult high school graduation rates, single-parent homes, and poverty levels can indicate the difficulty of educating students. Data indicate that as the percentage of children living in poverty in single-parent households with parents who did not graduate from high school increases, academic performance decreases. This analysis indicates that all three census variables affect student performance, and if any one were eliminated, performance could be expected to increase. Similarly, National Assessment of Educational Progress data from 42 states, when compared with census data, showed correlations between demographic characteristics and student performance. Changes should be made in how student test scores are compared and interpreted to provide a truer picture of student performance. (JPT)