It is argued that the reliability of archaeological inferences based upon artifactual assemblages is in part a function of the methods used to quantify the assemblages. Using ethnoarchaeological data, this proposition is tested by comparing associated artifactual assemblages with the presence or absence of specialists in individual households. Assemblages are quantified in terms of absolute frequency, diversity, and proportional frequency. The correlation between each measure and the presence or absence of specialists is assessed. The analysis reveals assemblage diversity to be one of the strongest and most consistent indicators of specialization and further reveals an inherent flaw characteristic of proportional frequencies.