Date of Submission


Date of Award


Institute Name (Publisher)

Indian Statistical Institute

Document Type

Doctoral Thesis

Degree Name

Doctor of Philosophy

Subject Name

Quantitative Economics


Theoretical Statistics and Mathematics Unit (TSMU-Kolkata)


Mukherjee, Diganta (ISRU-Kolkata; ISI)

Abstract (Summary of the Work)

Dual-record System (DRS) The problem of population size estimation is a very important administrative and statistical concern which includes a vast area of application in the fields of epidemiology, demography and official statistics. Federal agencies are generally interested to know the actual size (say, N ) of a specified population or any vital event that occurred in a specified area within a given time span. Census or civil registration system often fails to extract the true size of the population. The degree of inaccuracy depends on the actual size of the population, its diversity and of course, on the quality of the counting process. Any attempt to count all the individuals in a given moderate or large population is inevitably subject to error. As a remedy, the use of capture-recapture type experiment is being used for a long time. As per record, Laplace (1783 [60]) made the first implementation of such experiment in order to estimate the number of inhabitants in France. Thereafter, identical methods independently proposed by Petersen (1896 [73]) and Lincoln (1930 [63]) in order to estimate the size of an interest population became famous as Lincoln-Petersen estimate which is based only on one recapture operation after the first capture attempt. Individuals counted at the time of first capture are matched with the list of individuals prepared by the second attempt. In literature, this type of data structure with only two counting attempts covering the population is known as Dual-record System or simply, Dual System. A stabilized version of LincolnPetersen estimator developed by Chapman (1951 [24]) is still in use by numerous practitioners. Schnabel (1938 [81]) considered a multi-sample extension of the Lincoln-Petersen method, where each sample captured commencing from the second is examined for marked members and then every member of the sample is given another mark before being returned to the population. Later this multi-sample extension became very popular, especially for wildlifepopulations and commonly known as capture-mark-recapture (CMR) or simply capturerecapture experiment. Most of the advanced statistical models are developed after 1950’s in order to efficiently describe several situations arising in the real world for dependency between capture and recapture operations, varying individual capture probabilities, etc. Excellent reviews by Seber (1982[82], 1986[83]), Otis et. al.(1978[71]) and Chao (2001[23]) are available on capture-recapture theory. Thus, Dual-record System (DRS), which is particularly planned for human population, is technically very close to capture-recapture system with only two sampling occasions.In modern era, an application of the capture-recapture method is made in order to measure the extent of registration for vital events (ChandraSekar and Deming 1949 [17]; Ayhan, 2000 [1]) in the form of DRS. . Apart from the estimation of the size of a general population or vital events, this method has an extensive use for population growth estimation (Marks et al., 1974 [66]; Krotki, 1978 [59]), illusive or hard-to-count populations (Jibasen et al., 2012 [58]; Dreyfus et al., 2014 [33]) and also in several epidemiological applications including underascertainment in traditional epidemiological surveillance (Chao et al., 2001 [23]). Dual Record System Estimation is also used for the application to the problem of estimating undercount in census (Wolter, 1986 [102]; Cressie, 1989 [31]; Ayhan and Ekni, 2003 [2]; Elliot and Little, 2005 [34]; Watcher, 2008 [97]).


ProQuest Collection ID:

Control Number


Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.


Included in

Mathematics Commons