Shovel-Test Sampling as a Site Discovery Technique: a Case Study from Michigan

Shovel-test sampling is an increasingly common regional survey technique in archaeology. Recently, its effectiveness in discovering sites has been studied, but the paucity of empirical data has limited progress in this endeavor. In this paper, results of a survey in central Michigan are used to compare the effectiveness of shovel-test sampling and other survey techniques. Various techniques suitable to different land-use classes were employed. The results indicate that shovel-test sampling as generally applied performs poorly as a site discovery technique. Using control data from surface-collected sites, a simulation analysis shows the nature and degree of bias in results obtained by the technique. When properly employed, and with awareness of its serious shortcomings, shovel-test sampling can be useful. Its effectiveness, however, is limited and practical alternatives to its use must be considered.