Metric Data in Archaeology: A Study of Intra-Analyst and Inter-Analyst Variation

Metric data are regularly presented, analyzed, and compared. Despite acknowledgment that metric data can vary both when collected by one observer and when collected by multiple observers, few studies of these sources of variation in archaeological metric data have been undertaken. Intra-observer and inter-observer measurement errors are examined across four dimensions of 23 modern bighorn sheep (Ovis canadensis) astragali and five dimensions of 30 specimens of stone projectile point representing 4 culture-historical (temporal) types. Statistical and graphical analyses indicate that measuring dimensions of the same specimens multiple times facilitates determination of dimensions that can be readily and reliably measured and serves to screen data for data recording errors and for dimensions that may be subject to high levels of intra-observer and inter-observer variation.