Pointing out difficulties with a substantive interpretation of empirical data when using only traditional methods of data analysis such as crosstabulation of two or at most three variables, two recently suggested models - HARDER’s DO and GOODMAN’s ECTA - are briefly outlined. These two are combined in a new approach that has been developped by the author while analyzing data from two empirical studies. Some practical advice is included to help the average researcher with implementing this way of data analysis at his local computer installation.