You are working with the text-only light edition of "H.Lohninger: Teach/Me Data Analysis, Springer-Verlag, Berlin-New York-Tokyo, 1999. ISBN 3-540-14743-8". Click here for further information.

Exploratory Data Analysis

Exploratory data analysis (EDA) can be seen in contrast to traditional hypothesis testing. While the testing of hypotheses always requires an a priori assumption (or hypothesis) about the data (e.g. "There is a difference in the life expectation between smokers and non-smokers"), exploratory data analysis is not based on any a priori assumptions. Any method can be used to identify systematic relations between the variables.

In a typical exploratory data analysis all variables are taken into account using both graphical (e.g. scatter plots) and formal methods (e.g. principal component analysis) to search for systematic patterns.


Last Update: 2005-Jšn-25