Latent Class Analysis for Estimating an Unknown Population Size – with Application to Censuses

Abstract Estimation of the unknown population size using capture-recapture techniques relies on key assumption that capture probabilities are homogeneous across individuals in population. This is usually accomplished via post-stratification by some covariates believed to influence individual catchability. Another issue arises estimation from data collected multiple sources list dependence, where an individual’s catchability one related another list. The earlier models for heavily relied upon independence. However, there methods available can adjust estimates account dependence among lists. In this article, we propose use latent class analysis through log-linear modelling estimate presence both heterogeneity and dependence. proposed approach illustrated 1988 US census dress rehearsal.