Sampling and collector biases as taphonomic filters: an overview.

Abstract

Sampling (or sample) bias is a widespread concern in scientific research, across several disciplines. The concept of sampling bias originated in statistical studies. The consequence of a biased sample is that scientists will conclude about a population different from their target. In paleontology, sampling bias is typically related to fieldwork context. Human factors, known as sullegic (e.g. collection method, historic resampling) and trephic (transport, and curatorial processes) factors can generate bias. Other factor is the ugly fossil syndrome (i.e. choosing based on completeness of the specimens, or according to the researcher interest). Thus, sampling implies information loss. Biased samples add artificial results and can be considered an additional taphonomic filter. Therefore, sampling bias and the collector role and choices are frequently linked and almost indistinguishable. Compared to the treatment of this topic in other research fields, little related discussion has been held in vertebrate paleontology, especially regarding what happens at the interface between the biosphere, lithosphere, and anthroposphere, and during the transition between the anthroposphere and the patrisphere (museums). Numerous questions still arise. As a community, we must pay attention, to minimize the loss of information, from field activities to cataloging.