Appropriate statistics for ordinal level data: should we really be using t-test and Cohen’s d for evaluating group differences on the NSSE and other surveys?
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it