chapter  17
Selected Issues in Developmental Assessment
Pages 11

The extent to which we rely on measures to develop an interpretation of empirical phenomena should underscore heavily the necessity that measures be valid for the various uses to which they are put. Assessing changes-or differences that are interpreted as implying changes, as is frequently done in studying developmental phenomena-usually requires that one proceed as though what is being measured is conceptually the same or equivalent each time, or in each group, in order for comparisons to be meaningful. Or, just as the layman admonishes that one should not add oranges and bananas, neither should one subtract one from the other in computing changes or differences.