[Psychiatric and clinico-psychologic epidemiology: critique of methods and systematization approach]

The development of typical epidemiological questions is discussed. Possible definitions und fields of applications for simple status-diagnostics as well as for epidemiological longitudinal research are presented. Examples for the application of complex mathematical models like LOGIT- or LISREL-analysis are presented. Insufficient methodological systematisation and problems in comparing studies of different origin (i.e. for metaanalysis) are identified as problems for the future development of epidemiology. A classification system of epidemiological methods is presented that includes content aspects as well as methodological aspects on four levels. These are: the demand of the study (descriptive vs. inferential), the time dimension (status diagnostics vs. longitudinal research), the content level of operationalisation (micro- vs. macrolevel) and the level of explanation (correlative vs. causal). Examples show that this system can be used for classification of existing studies as well as for the conceptualisation and optimization of planned studies.