Ordinal Data can be arranged in some inherent order but differences between data
values either cant be
obtained or are meaningless. Subtracting data values is impossible or doesnt
result in something that means anything.
Examples:
Interval Data can be arranged in some inherent order, and meaningful differences
can be found between
data values. (It does make sense when you subtract data values.) There is no
inherent starting point to the data set (like 0) and ratios are meaningless. (You
cant use phrases like twice as or one-fourth as or three times as because
they dont mean anything).
Examples:
Ratio Data can be arranged in some inherent order, meaningful differences between
data values exist,
there is an inherent starting point to data values (0 would be pretty common),
and ratios between data values are meaningful. (It may be easier to think
about whether data has a starting point than whether ratios are meaningful. If
you are sure there is a starting point, then data is measured at the ratio level. If
you are sure there is not a starting point, then data cannot be at the ratio level.)
Example: