The power of a procedure for detecting mixture distributions in laterality data

The results of a Monte Carlo simulation of the power of an iterative version of a Maximum Likelihood procedure for the detection of mixture distributions are described. In general the power of such techniques is low, and for mixtures typical of those found in laterality research a sample size of about 150 or 300 is required to have a 50% or 90% probability of detecting the presence of a mixture.