Perhaps a quick and simple question.
If you have BMI scores in a data set, are they considered continuous/scale?
I have seen BMI given as an example of categorical/ordinal data in textbooks, but it would seem to me that it can only be considered categorical/ordinal once a new variable has been computed from this with values for the categories (i.e. 0: </= 18.4, 1: 18.5-24.9, 2: 25.0-29.9 ...) - is this correct? Should I therefore be treating BMI as continuous data unless it has been transformed into a new categorical variable as described above?
Any help would be great.