Blog roll
R-bloggers, Debian Weekly![[Valid RSS]](valid-rss-rogers.png "Validate my RSS feed")
Last modified:
oktober 17, 2019
Classical QCA operates on boolean variables, ie variables that are either TRUE or FALSE. Fuzzy Set QCA operates on continous variables, that are transformed into an ordered factor. This document intends to explain the basics about a) the transformation from continous into ordered factor and b) the fuzzy set QCA on the ordered factor.
I use the R package QCA by Adrian Dusa and Alrik Thiem, and that package includes the function calibrate() that is useful for this transformation. calibrate() offers advandes methods to classify the raw data into a discrete set of values, but here I'll just explain the basics.
Graph 1 shows, for 192 countries, the factor scores on a latent variable "Quality of Government".
The mean is essentially zero, the standard deviation is 0.34, the mininum value -0.732 and the maximum value 0.63
Now, let's say we want to transform this continous variable into a ordered factor with three groups. We do not want to manually inspect the distribution and manually decide on thresholds, what we want is a generic algorithm that can be applied to any dataset with a continous variable.