A categorical variable has a measurement scale consisting of a set of categories. Categorical variables that have ordered categories are called ordinal (Agresti, 2002). They appear, for example, whenever the condition of a patient cannot be measured by a metric variable and has to be classified or rated as “critical”, “serious”, “fair”, or “good”. The measurements on ordered categorical scales can be ordered by size, but the scales lack any algebraic structure; that is, the distances between categories are unknown. Although a patient categorized as “fair” is more healthy than a patient categorized as “serious”, no numerical value describes how much more healthy that patient is.
KeywordsGlasgow Outcome Scale Ordinal Data Adaptive Test Global Null Hypothesis Intersection Hypothesis
Unable to display preview. Download preview PDF.