We test the thesis that a materialism-postmaterialism value dimension underlies the twelve-item materialism-postmaterialism index used in the World Values Survey (WVS). Extending previous research on the logic of the four-item materialism-postmaterialism index which revealed fundamentally random response patterns in the United States, we focus on the logical and empirical relationships between the three four-item indexes that comprise the twelve-item index. If a single value dimension underlies the three four-item indexes, we should find moderate to high correlations between the three four-item indexes. Further, a finding of low inter-index correlations does not mean that individuals are responding randomly or insincerely; they may be answering in contextually interpretable ways that are nonetheless not consistent with a materialism-postmaterialism dimension. We apply three main methods: (1) a priori analysis of the internal logic of the materialism-postmaterialism indexes; (2) empirical analysis of the inter-index correlations, and confirmatory factor analysis, for all WVS respondents and within forty-two WVS societies in the 1990-93 WVS; and (3) logistic regression testing whether the priority that people give to materialist value items reflects the levels of material and physical security in their society. We find that the correlations between the three four-item indexes are nearly random, and in three-fourths of the countries the indexes fail to satisfy minimum criteria of dimensionality. "Postmaterialism" as measured by these indexes is not a value dimension in most WVS countries. Nonetheless, the priority given to particular material value items (e.g., maintaining order or fighting rising prices) is congruent with the prevailing social and economic conditions in given societies. But the usefulness of the data as indicators of material conditions has been obscured in previous research by the imposition of the materialism-postmaterialism classification scheme.