McFarland  12/96


Measurement Scales 


S. S. Stevens types of scales along a single dimension.

  
0.  Not a scale if categories overlap.
        Ex:  Person whose parents are of two different ethnic
		groups.
    Not a scale if categories aren't exhaustive.
        Ex:  Protestant/Catholic/Jewish doesn't provide
		categories for people with other religions or no
		religion.

 
1.  Nominal Scale.
    Scale consists of categories that are mutually exclusive and 
		exhaustive.
	Fixes:  "None of the above", "n.e.c.", "Other", "Choose
		the one response closest to your opinion".


2.  Ordinal Scale.
    Meaningful order along a single dimension.
    Ex:  Many attitude items:  strongly agree, agree, not sure, 
		disagree, strongly disagree.  	

 
3.  Interval Scale.
    Meaningful unit of measurement.
    Ex:  Time, with different calendars using different zeros.

 
4.  Ratio Scale.
    Meaningful zero point.
    Ex:  Amount of money; years of schooling.

 
Other considerations beyond Stevens:
 
1.  Is there a meaningful upper anchor point for the scale?
    Ex:  1.0 ?  100% ?  212 degrees F ?
    Ex:  Total RDA (recommended daily allowance) for nutrients ?
Is the upper anchor point a maximum, or can it be exceeded?

 
2.  Are negative values meaningful?
    Ex:  Loss = -(Profit)
    Ex:  Decline = -(Increase)
If the subject is change, can it go either direction?

 
3.  Is it fundamentally not a single dimension?
    Ex:  color wheel
 
4.  Is it mostly on a single dimension, the exceptions being such 
as "Not Applicable" or "Other" or "More than one of the above"?
 
 
 
Recommendation:  Unless there are good reasons otherwise, use 
the highest level of measurement scale the data permit.  Age as
a numerical value on a ratio scale, not age group as an ordinal
scale category.  There aren't many important differences
between 48 and 46 year olds, but I am sure you are well aware
of many important differences between 18 and 16 year olds, so
you would not want to lump them together in a single 16-20 age
group.