In scientific research on creativity, there has been considerable debate concerning the criteria by which a production can be judged more or less creative, that is, about the definition of "creativity." The most frequent definition -- the "standard" definition -- incorporates the criteria of "novelty" and "value." However, other definitions, based on a single criterion or on more than two criteria, have also been proposed. Much of the discussion of this issue has been based on semantic analysis, a logical analysis of the concepts involved and the usefulness of the various proposed criteria. In this article, question of the necessary and sufficient criteria for defining creativity is approached from an empirical (i.e., psychometric) perspective. The studies that are examined here converge on the idea that the standard definition is not internally consistent, because its two proposed criteria (i.e., novelty and value) are largely independent. Moreover, judgments of the creativity of an object seem to be explained mainly by its novelty, which suggests the possible sufficiency of that criterion. These results are consistent with the intentional novelty definition proposed recently by Robert W. Weisberg.