A prescriptive framework is provided for exploratory data analysis (EDA) in quality improvement projects. The three steps of EDA described are display of the data, identification of salient features, and interpretation of salient features. Graphical display of data, Shewhart's assignable causes, the maximum entropy principle, abduction, and explanatory coherence are part of the resulting framework. Additionally, the roles of probabilistic reasoning and automatic statistical procedures are examined.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados