Exploratory data analysis (EDA) is a conceptual framework with a core set of ideas and values aimed at providing insight into data as it is presented to the working researcher (regardless of its origin), and to encourage understanding probabilistic and nonprobabilistic models in a way that guards against erroneous conclusions. Because this set of goals covers experimental and nonexperimental data, clean and messy data, and data in forms that may not be properly statistically modeled, Tukey distinguished these goals from the more specific probabilistic goals of traditional "statistics," which he referred to as "confirmatory data analysis" (CDA). Clearly these practice-based and pragmatic goals are well aligned with the needs of active researchers in the psychological community (Behrens, 1997a). Although an explicit account of EDA is slowly growing in the psychology literature, the influence of Tukey's principles are as far reaching as key works in philosophy of data analysis (Cohen, 1990, 1994; Nickerson, 2000), regression graphics (Cook & Weisberg, 1994), robustness studies (Wilcox, 1997, 2001), and computer graphics for statistical use (Scott, 1992; Wilkinson, 2005). Despite these influences, recent research regarding the training of psychological researchers suggests little advancement in psychologists' abilities to apply common techniques important to the EDA tradition, including dealing with nonlinearity, advanced graphics, or model diagnostics (Aiken, West, & Millsap 2008). Likewise, it is not unusual that authors of papers published in referred journals neglect detailed examination of data. To introduce the reader to EDA, the chapter is divided into four parts. First, the background, rationale, and philosophy of EDA are presented. Second, a brief tour of the EDA toolbox is presented. Third, computer software and future directions for EDA are discussed. The chapter ends with a summary and conclusion.