The three-parameter Weibull density is commonly used to model the dist
ribution of tree diameters in forest stands. We demonstrate, through l
ikelihood profiles, that maximum likelihood estimation is often inappr
opriate for data from young trees due to negative estimates of the loc
ation parameter. We suggest a Bayesian model and fit it, using the Gib
bs sampler, to three data sets. The latter model is easy to implement
and guarantees a positive estimate for the location parameter. We illu
strate some novel forms of model diagnostics, demonstrating that the B
ayesian model is appropriate for two of the data sets, while it is dub
ious for the third. A sampling-resampling method shows that the lack o
f fit of the model for the latter data set is due to the likelihood, a
nd not the prior specification.