The Practice of Social Research

Chapter Five.  Conceptualization, Operationalization, and Measurement

OPERATIONALIZATION CHOICES
    Range of Variation
    Variations Between the Extremes
    A Note on Dimensions
    Defining Variables and Attributes
    Levels of Measurement
    Single or Multiple Indicators
    Some Illustrations of Operationalization Choices
    Operationalization Goes On and On
 

    Operationalization is a step beyond conceptualization in which we specify precisely how we will measure or recognize our concepts in the course of making observations.  This section will touch on some of the many choices available in that regard.

    You often have a choice of range, whether to distinguish extremes or settle for more moderate distinctions.  Similarly, you can make fine or gross variations between the end points: e.g., you can ask people their precise age in years or settle for their decade (20s, 30s, etc.)  We'll return to the issue of dimensions and see the choice you have in that respect.  As you'll see, operationalization concludes the definitional problem initiated in conceptualization.

    There are several levels of measurement, reflecting the relationship among the attributes comprising a variable.  In nominal variables, the attributes are simply different (e.g., male and female).  Ordinal variables rank-order the attributes (e.g., Always, sometimes, never).  Interval variables add the quality of standardized units (e.g., Farenheit and Celsius temperatures), and ratio variables include a genuine zero (e.g., age).  While some variables--like gender--can only be measured at the nominal level, others--like age or income--offer you a choice.

    Sometimes we measure variables with single indicators, sometimes we combine several indicators into a single measure.