Capture-Recapture models when both sources have clustered observations

The model most commonly used in capture-recap-ture estimation of population sizes assumes that ind iv iduals are missed or captured in each source independently of a l l other indiv iduals in the population. However, th i s assumption may be inadequate for some studies: certain research designs would lead natura l ly to c luster ing of misses in the enumeration of a populat ion. This paper w i l l develop a model that w i l l describe the c luster ing of misses, i t w i l l describe biases inherent in the t rad i t iona l dual system estimator when the model involving c luster ing holds, and f ina l ly, the paper w i l l out l ine the use of the EM algorithm to estimate the to ta l population size