Exploratory Data Analysis

This chapter presents exploratory data analysis (EDA) as an approach for gaining understanding and insight about a particular dataset in order to support and validate statistical findings, as well as potentially generate, identify, and create new hypotheses based on patterns in data. Examples of EDA are provided and interpretations are discussed. EDA may be used at any stage in the data analysis process from cleaning through transformation and descriptive analysis, as well as using results from every stage. Discovery of patterns may inspire a new direction in intervention effectiveness research, as well as further supporting or validating existing projects. Data visualization skills are essential for individuals working with large datasets.