I. INTRODUCTION Once the researcher has conceptualized a model, perhaps graphically representing it in a circle-and-arrows diagram, he or she then faces the task of measuring each of the variables in the model. This process is called operationalizing variables. For every variable in the model ("concept" or "construct"), there must be at least one and preferably multiple specific measures ("indicators"). Thus, operationalization is the associating of indicator variables with conceptual variables. Many issues arise in this process. Even the seemingly trivial challenge of operationalizing the conceptual variable "Subject's gender" can run into problems. A seemingly sound survey item that calls for self-reporting by checking' 'male" or "female" will normally suffice, but in a large population, such an item will yield results different from the operational item, ' 'Were you born a genetic female?" To take a second example, the conceptual variable "income" may be operationalized in terms of family income or individual income, and in terms of cash income versus total value of benefits received in cash or in kind. The number of people below the poverty line will change significantly when in-kind goods and services such as food stamps or subsidized daycare are included in the measurement of "income."