Search

This data article describes how the Dig It Check It (DICI) model operates. The purpose of the DICI model is to assess the effectiveness of a specified subsurface sampling program in the detection of archaeological sites (Way and Tabrett, in press) [1]. Specifically, the aim of the model is to determine the inherent biases of the specified sampling program, i.e. what is the probability that sites of a certain size, density and density-distribution will be, or were, missed by the sampling program. A simulation is run which randomly places sites, the characteristics of which can be varied between runs, in a sampling area with the specified test-pit layout to determine the probability of overlap between the test-pits and sites with the selected characteristics. When overlap does occur, the site is recorded as intersected. The model then calculates the site density at that point and compares it with the test-pit size to determine whether the density is great enough for detection to also occur.

This article describes in detail how the Dig It, Design It (DIDI) simulation tool operates to design a subsurface landscape sampling strategy and predict its likely effectiveness. The purpose of the DIDI model is to help archaeologists develop statistically sound subsurface sampling programs that maximise the number of sites found while minimising the number of sampling units used. It has been unusual for archaeological test-pitting programs to be theoretically tested or statistically justified by simulation prior to implementation. Previous research by Kintigh (1988) and Krakker et al. (1983) established the statistical principles underlying the subsurface sampling of a rectangular survey area, and Kintigh pioneered the use of Monte Carlo simulations to test the effectiveness of these sampling strategies. DIDI provides an updated version of this simulation approach that has three key benefits over the original version: 1. It allows the user to model a larger range of possible archaeological conditions by providing an additional density distribution function (see below), and making the clustering parameter available with all distribution functions; 2. DIDI gives the user the option of filling the previously unavoidable gaps around the edges of the survey area with additional, suitably placed test-units, thereby increasing the detection rate of a sampling strategy; and 3. It is free to download and use and is compatible with modern operating systems.