The purpose of this app is to optimise the location of cut-off points in numeric data, such that the resulting binary data version of an independent variable is the best available predictor of a binary version of the dependent variable. These binary versions of data can then be used for crisp-set QCA and EvalC3 type analyses.
There are four analysis options at stage 2:
Maximum file size: 60 rows, 50 columns
Example of a correctly formatted CSV file:
ID,IV1,IV2,IV3,DV1,DV2,... 1,23,45,67,89,12 2,34,56,78,90,23 ...
(IV = Independent Variable, DV = Dependent Variable)
Classification Accuracy:
Balanced Accuracy:
F1 Score:
Precision (Sufficiency):
Recall (Necessity):
MCC:
Optimal Cut-off Point(s):