Path analysis

In statistics, the term path analysis describes a form of examining the dependencies between variables. As part of the path analysis, path models , i.e. H. theoretically derived models of causal relationships between variables, empirically verified. The path analysis is part of the causal analysis. The relevant causal factors must be identified even before the modeling can be started. This problem can be solved in different ways.

The correlation between two variables does not yet represent a causal relationship per se, but merely provides the "explanandum" for a scientific explanation (see Hempel-Oppenheim scheme ). In other words: the statistical correlation relationship in itself does not provide an explanation, but must itself be explained theoretically. For example: "Divorce rates" are "time" dependent. The independent variable "time" has yet to be interpreted sociologically in order for the explanation to be provided in full, for example: that the degree of urbanization has changed over time.

Demarcation

Path analysis is a form of multiple regression analysis oriented towards causal relationships, but can also be viewed as a special case of a structural equation model in which only individual indicators are used for the respective variables of the causal model. It is therefore a structural equation model in which the structural model is present but the measurement model is dispensed with.

history

The geneticist Sewall Wright developed the path analysis around 1918 and described it in detail in 1920. Outside of genetics , path analysis is used in particular in sociology and econometrics . The original claim of this statistical method to show causal relationships was discussed intensively from the beginning, but has now receded somewhat into the background.

literature

• Jürgen Bortz & Christof Schuster: Path analysis . In: Statistics for human and social scientists . 7th edition. Springer-Verlag, Berlin [a. a.] 2010, p. 435-452 , doi : 10.1007 / 978-3-642-12770-0_24 .
• Sewall Wright: The Method of Path Coefficients . In: Annals of Mathematical Statistics . tape 5 , no. 3 , 1934, pp. 161-215 , doi : 10.1214 / aoms / 1177732676 , JSTOR : 2957502 .

Individual evidence

1. ↑ Chong Ho Yu: Causation in Quantitative Research Methodologies from Path Modeling, SEM to TETRAD. Theory & Science (2007). ISSN  1527-5558 .
2. ↑ Hartmut Esser : Sociology. General basics. Campus Verlag Frankfurt / New York 1993. ISBN 3-593-34960-4 . P. 89
3. ^ Wright, S. (1921) Correlation and causation . J. Agricultural Research, 20, 557-585
4. ^ Dodge, Y. (2003) The Oxford Dictionary of Statistical Terms. OUP. ISBN 0-19-920613-9
5. ^ Daniel J. Denis, Joanna Legerski: Causal Modeling and the Origins of Path Analysis Theory & Science (2006). ISSN  1527-5558