What is it about?
Some researchers argue that QCA solutions change "a lot" on tiny variations of the input data: some of the values being replaced, or the value of the outcome changed, leading to "dramatically" different solutions. This article tackles the sensitivity diagnostics, by employing a combinatorial approach to calculate highly precise retention (stability) probabilities of the baseline solutions: it shows the proportion of situations where the QCA solution is the same (inversely, the proportion of situations where it changes) by either corrupting some values, or deleting cases altogether.
Featured Image
Why is it important?
There were several previous attempts to calculate sensitivity measures, some arbitrarily setting values that would lead to certain outcomes, while other employing imprecise calculation methods. This article is the first to present exact retention probabilities under various scenarios.
Perspectives
The article looks more complicated than it actually is. Essentially, it is about a relative frequency of retaining baseline solutions, and there is an R function in the replication files, which I wrote especially for this article, that does all the heavy work. The function, called "retention()", is already implemented in the QCA package in R: https://cran.r-project.org/web/packages/QCA/
Prof.Dr. Adrian Dușa
University of Bucharest
Read the Original
This page is a summary of: Enhancing Sensitivity Diagnostics for Qualitative Comparative Analysis: A Combinatorial Approach, Political Analysis, January 2016, Cambridge University Press,
DOI: 10.1093/pan/mpv028.
You can read the full text:
Contributors
The following have contributed to this page
