Abstract
Shotwell's method of paleoecological reconstruction attempts to define community membership of mammals represented within paleontological assemblages. Analysis of the measure of community membership upon which the method is based (relative skeletal completeness) shows this measure to be a function of sample size. A method for controlling for the effects of sample size upon measures of relative skeletal completeness is needed. A method is suggested which employs ratios of numbers of individuals to numbers of specimens, and a specific correction of Shotwell's method is proposed.