Non-Parametric Statistics
Categories: Metrics
Non-parametric statistics are values we calculate that do not require that we’ve met certain conditions before we even begin the calculation. This is in contrast to calculations like ANOVA or confidence intervals or hypothesis tests (and loads of other statistical processes), which make us jump through a load of conditional hoops before we can even start to run the numbers.
Non-parametric statistics are typically things like the number of stars out of five that a hotel gets rated on a travel website, or the average customer satisfaction score for a business, or even non-numerical values, like “strongly agree/agree/disagree/strongly disagree.” Because we don’t have to ensure a bunch of conditions were met before we perform the calculations, non-parametric statistics often contain a greater degree of uncertainty than parametric ones, where we jumped through those conditional hoops.