Reframes population size estimation as a Bayesian regression problem to correct for heterogeneous detection probabilities in cultural samples.
Explores Zelterman's robust alternative to Chao1 — an estimator less sensitive to heterogeneity violations, tested on cultural and literary data.