Educational research, unfortunately, often focuses on finding statistical differences between overall means or averages. Media reports of research routinely present those differences and little else. This paper discusses the importance of considering the spread of the data in addition to the center and how this is relevant to research focused on rural schools. An example from one Kentucky county shows how, in the case of a large group of fourth-grade students with higher average test scores than a much smaller group, the difference between mean scores is misleading and draws attention away from the considerable overlap in the distributions of the two groups' scores. In addition, while the larger group had a higher mean score, it also had many more low-performing students than the smaller group. This county-wide data was broken down further to show the distributions of Group-1 and Group-2 scores in each of the county's six elementary schools. The patterns differed markedly among the schools, suggesting many questions for research. In another example, variance decomposition is used to portray mathematics achievement scores for Japanese and U.S. students in terms of whether the variation is between students, between classrooms, or between schools. Comparisons of achievement status versus achievement growth are also considered. The relevance of these statistical issues to rural education research are discussed in terms of research on small schools, small classes, and the relationships between rural student backgrounds and achievement. (SV)