Bayesian analysis of multi-source data

We present a simple method, based on Bayes' theorem, to fit binned data to one or more multi-source models. Assuming a Poisson probability for the count in each bin we can eliminate exactly the nuisance parameters from the likelihood function and arrive at a formula that can be broadly applied. We illustrate the method by showing how it can be used to estimate the top quark mass. (C) 1997 Published by Elsevier Science B.V.