The Practice of Social Research

Chapter Five.  Conceptualization, Operationalization, and Measurement

CONCEPTUALIZATION
    Indicators and Dimensions
    The Interchangeability of Indicators
    Real, Nominal, and Operational Definitions
    Creating Conceptual Order
    An Example of Conceptualization: The Concept of Anomie

    Concepts are mental images with labels like "chair," "female," "social class," and "grade-point average."  The conceptualization process refines our concepts into attributes and variables, introduced in Chapter One.  As we refine what we mean by a concept such as "social class," for example, we also consider possible indicators of it (observations that would give evidence of low or high class) and different dimensions or aspects of the concept.  Thus social class can be a matter of economics, prestige, power, etc.--different dimensions of the concept.

    Even if we can't agree on the proper indicators for a given variable, we may be able to pursue it in an explanatory analysis.  For example, if women are more religious than men on all our indicators of religiosity, we are probably safe in saying they are more religious in a general sense--even though we can't agree on what "religious" means.  This is the interchangability of indicators.

    Most modern researchers reject the notion of real definitions, that the terms we use have inherent meanings.  The idea of nominal definitions, by contrast, views labels and definitons as linguistic and intellectual agreements.  Operational definitions specfy the manner in which we will measure the variables that interest us.  So if I tell you that I will measure "social class" by asking people for their self-perceptions, you may disagree with that technique but you will at least know what my definition is.

    As we'll see, concepts are seductive.  Once we have created concepts like "incorigible," "executive burnout," or "model student," it's easy to begin believing those terms represent something real in nature--rather than linguistic conventions.