Ihre Browserversion ist veraltet. Wir empfehlen, Ihren Browser auf die neueste Version zu aktualisieren.

Symposium "Qualitative Comparative Analysis in Critical Perspective"

Posted 27/11/2014

The journal Sociological Methdology, edited by Tim Futing Liao, enjoys a reputation for publishing high-quality work in the field of social research methods.

This is why it is all the more unfortunate that the symposium "Qualitative Comparative Analysis in Critical Perspective" by Samuel R. Lucas and Alisa Szatrowski, published in Sociological Methodology, 44(1), marks an unprecedented low point in the increasingly heated debate on QCA, verbally and methodologically.

The lack of proper documentation causes many pages to be wasted on guesses by the commentators about data errors, errors in the analysis and replication problems. Neither the results of the first example in the study, which analyzes the space shuttle data, nor those of the first study with simulated data, are replicable. Worse still from the perspective of Lucas & Szatrowski, QCA does find the correct solution where the authors claim it does not.

However, the knee-jerk rejections of simulation studies by some commentators are equally unfortunate. Lucas & Szatrowski as well as Fiss, Marx, Rihoux, and Vaisey are right when arguing that a key way to establish the validity of a method is to use simulated data. If a method fails even with ideal simulated data, for which all external disturbance factors such as omitted variables bias can be eliminated, how can the method be expected to be able to handle empirical data?! I completely understand Lucas' frustration at this issue.

Simulations are powerful tools, but they must be handled appropriately. In this connection, it is of course pointless to let QCA try to recover a given regression function or parts of this function, as Lucas does. No regression method would in turn be able to recover any given Boolean expression. The tests simply do not match the method to be assessed and are therefore invalid. As a result, the core argument of Lucas & Szatrowski, that QCA is unsafe to use, immediately collapses for the simple reason of a gross ignorance of the mathematical foundations of Boolean and fuzzy algebra.

Overall, then, what to make of this symposium? Apart from Lucas & Szatrowski's call at proponents of QCA to prove that their method works by using simulation studies, I found nothing to commend it for. Quite on the contrary, instead of the step forward that I was hoping for, two steps back have been made.

Fields marked with * are required.