The focus of this study was on the relationship between visual and statistical analyses of time series data and the degree to which characteristics of the data influenced this relationship. A total of 52 teachers took part in evaluating a series of graphs having prespecified characteristics. The independent variables manipulated included: slope, variability, training, and aimline/decision rules. Generally, the influence of these variables was significant and in the predicted direction. However, the overall level of relationship between the two analytic procedures was modest. The implications of this research are discussed in terms of the manner in which visual analysis is conducted and the procedures needed to establish statistical conclusion validity. (Author)