INDEX CONSTRUCTION
    Item Selection
    Examination of Empirical Relationships
    Index Scoring
    Handling Missing Data
    Index Validation
    The Status of Women: An Illustration of Index Construction

    This section of the chapter takes you step-by-step through the construction of an index.  First, you must select a set of data items which you believe--on face value--to reflect the variable being measured: several questionnaire items that reflect prejudice, for example.

    Next, you examine bivariate and multivariate relations among the items, seeing if they are empirically correlated with one another.  If two items are not related to each other, there is a good chance they do not reflect the same variable.  In this fashion, you winnow away bad indicators in favor of good ones.

    Having selected a set of indicators of the variable, you must assign points to each.  Usually, you will simply give subjects one point for each of the indicators and total their scores.  If one indicator seems to reflect the variable more extremely than others (e.g., represents a high level of prejudice), you could assign more points to that indicator.

    Once you have created as index, you validate it as a measure of the variable by seeing whether it correlates with other data items in ways it should.  Does it predict responses to other items reflecting prejudice, for example?