Describing Data and Testing Hypotheses Quiz

See Variable Types. Option B is correct because severity here is an ordinal variable: its categories have a meaningful order but unequal or unknown spacing. Mild, moderate and severe clearly rank from least to most, but nothing says the step from mild to moderate is the same size as the step from moderate to severe. Typing the labels as 0, 1 and 2 records the order so that a spreadsheet can sort and count them, but it does not turn the labels into continuous measurements (values that can take any number in a range, like height) or make their mean automatically meaningful. The variable type, not the digits in the spreadsheet column, decides which summaries are valid. For ordinal data a sensible summary is the number or proportion in each category, or the median, and rank-based methods are often more appropriate than methods that assume equal intervals. Compare this with a nominal variable such as blood group, which has no order at all, and with a continuous variable such as serum sodium, where the spacing between values really is equal.

• Severity: Mild, moderate and severe have a meaningful order but unequal or unknown spacing, so it is an ordinal variable.
• Numeric codes: 0, 1 and 2 record the order only and do not make the data continuous or the mean meaningful.
• Valid summaries: The number or proportion in each category, or the median.