Volume 57 - Number 2 - Summer 1991

N. Methods of Research and Analysis Including Models

Studies concerned with demographic methods and with methods from other disciplines that have been applied to demographic data as a whole. Includes mathematical demography and studies on methods of estimation and indirect estimation. Methodological studies and models concerned with one demographic variable, such as migration, are coded under the category concerned with that topic and cross-referenced to this heading. Studies on models used to investigate relationships between demographic variables and for the analysis of empirical data are also coded under this heading.

57:20738 Biaye, Mady. Being in a state, the succession of states, and the length of stay in each state in event history analysis: a review of where we stand in the analysis of mortality. [La presence des etats, leur ordre de succession et la duree de sejour dans chacun d'eux dans l'analyse des histoires de vie: un essai de prise en compte dans l'analyse de la mortalite.] Institut de Demographie Working Paper, No. 154, ISBN 2-87209-131-9. 1991. 27 pp. Universite Catholique de Louvain, Institut de Demographie: Louvain-la-Neuve, Belgium. In Fre.
This paper is an attempt to adapt the concepts of the order of succession of states and duration of stay in each state to event history analysis by a process of adapting some existing analytic methods. The author notes that not only is the value of the new models dependent on the correctness of the hypotheses on which they are based, but the results obtained from their application depend on the quality of the data available.
Correspondence: Universite Catholique de Louvain, Institut de Demographie, Place Montesquieu 1, Boite 17, 1348 Louvain-la-Neuve, Belgium. Location: Princeton University Library (SPR).

57:20739 Boozer, Michael A.; Guinnane, Timothy W. The consistency of reported marital status and information drawn from own-child checks in the 1910 census of the United States. OPR Working Paper Series, No. 91-4, Dec 1990. 9, [15] pp. Princeton University, Office of Population Research [OPR]: Princeton, New Jersey. In Eng.
"The 1990 Federal Census of the United States did not ask married individuals whether they had previously been married. Several researchers have used 'own-child checks' to substitute for direct information on remarriage in selecting couples for marital fertility analysis from the Public Use Sample of the U.S. Federal Census of 1900. This note compares the results of such checks to direct information on remarriage in the Public Use Sample of the U.S. Census of 1910. The 1910 census did ask married persons whether they had been married previously. Comparison of the direct and indirect information on remarriage shows that the checks detect fewer than two-thirds of wives who report they are remarried. On the other hand, parity distributions for women who say they are remarried but are not identified as such by the own-child checks are very similar to those for the population of remarried women as a whole. The own-child checks are a poor method for the study of remarriage per se, but the checks perform remarkably well as part of fertility analysis of the 1900 Public Use Sample."
Correspondence: Princeton University, Office of Population Research, Working Paper Series, 21 Prospect Avenue, Princeton, NJ 08544-2091. Location: Princeton University Library (SPR).

57:20740 Brant, L. J. Methodological issues in the epidemiology of risk assessment and mortality. Collegium Antropologicum, Vol. 14, No. 1, Jun 1990. 107-22 pp. Zagreb, Yugoslavia. In Eng. with sum. in Scr.
"Risk factor analyses are widely used in epidemiological studies of morbidity and mortality. This paper discusses different models that are used for such studies and addresses numerous issues that arise in the subsequent analyses. Methodological issues are illustrated by examining the relationship between relative body weight and mortality in a cohort of [U.S.] males participating in a longitudinal study of normal aging."
Correspondence: L. J. Brant, National Institutes of Health, National Institute on Aging, Gerontology Research Center, 4940 Eastern Avenue, Baltimore, MD 21224. Location: Stanford University Libraries, Stanford, CA.

57:20741 Denton, Frank T.; Feaver, Christine H.; Spencer, Byron G. The MEDS projection and simulation system. QSEP Research Report, No. 263, Jun 1990. 27 pp. McMaster University, Faculty of Social Sciences, Program for Quantitative Studies in Economics and Population: Hamilton, Canada. In Eng.
"MEDS is an acronym for Models of the Economic-Demographic System, a system that consists of a series of interrelated models designed for use in projecting or simulating the development of the Canadian population and economy over the longer term. MEDS now includes four models, and each of these is discussed briefly. The models are designed for easy use on standard DOS-based microcomputers. Recent changes incorporated in them are discussed, and plans for further development of MEDS are noted."
For earlier reports on the MEDS models, including a machine-readable data file, see 56:20768 and 30074; for a related study by the same authors, see elsewhere in this issue.
Correspondence: McMaster University, Faculty of Social Sciences, Program for Quantitative Studies in Economics and Population, Hamilton, Ontario L8S 4M4, Canada. Location: Princeton University Library (SPR).

57:20742 Denton, Frank T.; Feaver, Christine H.; Spencer, Byron G. The MEDS system of models: an instrument for economic-demographic simulation and projection. QSEP Research Report, No. 269, Oct 1990. 33 pp. McMaster University, Faculty of Social Sciences, Program for Quantitative Studies in Economics and Population: Hamilton, Canada. In Eng.
"MEDS (Models of the Economic-Demographic System) is a series of interrelated user-friendly computer models designed for use in projecting or simulating the development of Canadian population and economy over the longer term. There are currently four MEDS models, each designed for easy use on standard DOS-based microcomputers. The models are discussed briefly, and some recent extensions relating to life cycle behaviour, human capital calculations, and education and training expenditures are illustrated."
For earlier reports on the MEDS models, including a machine-readable data file, see 56:20768 and 30074; for a related study by the same authors, see elsewhere in this issue.
Correspondence: McMaster University, Faculty of Social Sciences, Program for Quantitative Studies in Economics and Population, Hamilton, Ontario L8S 4M4, Canada. Location: Princeton University Library (SPR).

57:20743 Feeney, Griffith. Untilting age distributions: a transformation for graphical analysis. Asian and Pacific Population Forum, Vol. 4, No. 3, Fall 1990. 13-20 pp. Honolulu, Hawaii. In Eng.
"This article presents a new approach to the plotting of age distribution data. 'Untilting' is a way of transforming data that vary systematically from very high to very low values so as to show local variation more clearly. The article derives an untilting transformation from the formal structure of age distributions. The transformation turns out to be closely related to two familiar demographic techniques, reverse-survival estimation of births and birth rates, and comparison of observed with stable age distributions. The ideas are illustrated by application to age distributions from the 1979 and 1989 censuses of Vietnam."
Correspondence: G. Feeney, East-West Center, East-West Population Institute, 1777 East-West Road, Honolulu, HI 96848. Location: Princeton University Library (SPR).

57:20744 Ferguson, Brian S. Optimal population control with uncertain output. Journal of Population Economics, Vol. 3, No. 4, 1990. 291-302 pp. New York, New York/Berlin, Germany, Federal Republic of. In Eng.
"This note has shown how the techniques of stochastic control theory can be used to analyze the impact of uncertainty about non-labour inputs on optimal population control expenditure. While we have chosen very simple structure, it has allowed us to isolate the channels through which uncertainty affects the optimal plan, in this case causing expected expenditure on control to rise more slowly or fall more rapidly than its deterministic counterpart. Given the long-term implications of changes in population growth at any instant, and the stochastic nature of the environment in which control decisions are made, it is likely that useful insights could be obtained by extending the stochastic control approach to more complicated population control structures."
Correspondence: B. S. Ferguson, University of Guelph, Department of Economics, Guelph, Ontario N1G 2W1, Canada. Location: Princeton University Library (SPR).

57:20745 Gray, Alan. Analysis of components of demographic change. Mathematical Population Studies, Vol. 3, No. 1, 1991. 21-38 pp. Reading, England. In Eng.
"There exist a number of approaches to analysis of the components of difference between demographic rates using standardization techniques. A generalization of some of these methods, for an arbitrary number of components and an arbitrary analytical function of them, is shown to be the theoretically best choice when the analytical function is linear in each of the components and when linear change is assumed for each of the components. Also, a very close approximation to an exact decomposition for the most general case can be obtained when dealing with change over time and time series data are available."
Correspondence: A. Gray, Australian National University, National Centre for Epidemiology and Population Health, GPO Box 4, Canberra ACT 2601, Australia. Location: Princeton University Library (SPR).

57:20746 Greenhalgh, David. An epidemic model with a density-dependent death rate. IMA Journal of Mathematics Applied in Medicine and Biology, Vol. 7, No. 1, 1990. 1-26 pp. Oxford, England. In Eng.
"This paper deals with a mathematical model for a disease where the death rate depends on the number of people in the population. This sort of model would be suitable for diseases in developing countries....We also assume that the population under consideration is regulated by the disease so that there is a mortality induced by the disease. We use a compartmental model and perform an equilibrium and stability analysis to find that there is a threshold condition. If the threshold is exceeded, then there is a unique equilibrium with disease present which is locally stable to small perturbations."
Correspondence: D. Greenhalgh, University of Strathclyde, Department of Mathematics, Livingstone Tower, 26 Richmond Street, Glasgow G1 1XH, Scotland. Location: Princeton University Library (SPR).

57:20747 Khalifa, Mona. The loglinear model as a tool for the analysis of demographic data. Egyptian Population and Family Planning Review, Vol. 24, No. 1, Jun 1990. 17-41 pp. Giza, Egypt. In Eng.
The author discusses the use of loglinear models in demographic analysis, particularly with regard to the analysis of contingency tables. Their use is illustrated using data from the 1980 World Fertility Survey for Sudan to analyze the determinants of fifth-birth intervals. The results show "the ability of the technique to improve our understanding of the mechanism of fertility differentials using individual--instead of aggregate--data. It has shown how each variable affects the process when all other variables are held constant."
Correspondence: M. Khalifa, Cairo University, Department of Statistics, POB 1055, Khartoum, Sudan. Location: Princeton University Library (SPR).

57:20748 Marshall, Roger J. Mapping disease and mortality rates using empirical Bayes estimators. Journal of the Royal Statistical Society, Series C: Applied Statistics, Vol. 40, No. 2, 1991. 283-94 pp. London, England. In Eng.
"Methods for estimating regional mortality and disease rates, with a view to mapping disease, are discussed. A new empirical Bayes estimator, with parameters simply estimated by moments, is proposed and compared with iterative alternatives suggested by Clayton and Kaldor." The author develops a local shrinkage estimator in which a crude disease rate is shrunk toward a local, neighborhood rate. The estimators are compared using simulations and an empirical example based on infant mortality data for Auckland, New Zealand.
Correspondence: R. J. Marshall, University of Auckland, School of Medicine, Department of Community Health, Private Bag, Auckland, New Zealand. Location: Princeton University Library (PR).

57:20749 McCall, Brian P. Testing the proportional hazards assumption in the presence of unmeasured heterogeneity: an application to the unemployment durations of displaced workers. Center for Population Analysis and Policy Research Report, No. 91-03-2, Mar 1991. 38 pp. University of Minnesota, Humphrey Institute of Public Affairs, Center for Population Analysis and Policy: Minneapolis, Minnesota. In Eng.
"This paper develops tests of the proportional hazards assumption when duration data is grouped and when some heterogeneity, which influences the duration of an event, is unmeasured....These tests are applied in studying the determinants of unemployment durations among displaced workers. Unemployment duration data is derived from the 1986 and 1988 [U.S.] Current Population Survey's Displaced Worker Supplement. The proportional hazards assumption is rejected for the full sample but cannot be rejected for a subsample of individuals who are eligible for unemployment insurance benefits."
Correspondence: University of Minnesota, Humphrey Institute of Public Affairs, Center for Population Analysis and Policy, 301 19th Avenue South, Room 230, Minneapolis, MN 55455. Location: Princeton University Library (SPR).

57:20750 Mitra, S. About Holland's "On the adequacy of Mitra's model of the life table: a technical note" Demography India, Vol. 17, No. 2, Jul-Dec 1988. 329-30 pp. Delhi, India. In Eng.
This is a response to a critique by Bart Holland of the author's 1983 article on modeling life table functions.
For the article by Holland, published in 1987, see 54:20209.
Correspondence: S. Mitra, Emory University, Atlanta, GA 30322. Location: Princeton University Library (SPR).

57:20751 Poltavtsev, Yu. G. A problem of modeling population growth. [Problema modelirovaniya vosproizvodstva naseleniya.] Demograficheskie Issledovaniya, Vol. 14, 1990. 34-43 pp. Kiev, USSR. In Rus. with sum. in Eng; Ukr.
"An attempt is made to simulate [population] reproduction which is considered as a [combination] of labour and demographic activity."
Location: Princeton University Library (SPR).

57:20752 Potter, Joseph E. Demographic approaches to survival and reproduction: parallels between the mortality transition and the fertility transition. Texas Population Research Center Paper, No. 12.07, 1990-1991. 24 pp. University of Texas, Texas Population Research Center: Austin, Texas. In Eng.
"In this essay, after offering a brief and rough caricature of recent social science research on fertility, I will look at what appear to be the reasons why social scientists have considered mortality and fertility to be quite different phenomena, subject to quite different sorts of inquiries. The discussion will proceed from there to take up the numerous similarities between the two fields of research. The thrust of the argument is that, at least among demographers, the 'social' and 'behavioral' nature of health and mortality has been severely underestimated heretofore, that there is considerable overlap and complementarity between fertility research and mortality research, and that most of the customary reasons for thinking that mortality research is inherently less social and behavioral than fertility research are not 'good reasons'." Attention is given to the methodological implications of this approach. The geographical focus is on developing countries.
Correspondence: University of Texas, Texas Population Research Center, Main 1800, Austin, TX 78712. Location: Princeton University Library (SPR).

57:20753 Rees, Philip. Old model faces new challenges: a review of the state of the art in multistate population modelling. School of Geography Working Paper, No. 531, Sep 1989. iii, 49 pp. University of Leeds, School of Geography: Leeds, England. In Eng.
"The paper reviews the state-of-the-art in multistate population modelling. After a thumbnail sketch of the projection version of the model, a series of challenges to the model structure and utility are considered....The multistate model's cousin, the dynamic microsimulation model, is then analyzed as a response to the challenge to enrich model outputs. Finally, the paper discusses two new challenges to...multistate modellers: to deal with the AIDS pandemic, and to respond to the [general information system] challenge."
Correspondence: University of Leeds, School of Geography, Leeds LS2 9JT, England. Location: Princeton University Library (SPR).

57:20754 Sugarev, Zdravko. Some generalizations of the net coefficient of population reproduction. [Nyakoi obobsteniya na neto-koefitsienta za vazproizvodstvoto na naselenieto.] Naselenie, Vol. 7, No. 4, 1989. 40-51 pp. Sofia, Bulgaria. In Bul. with sum. in Eng; Rus.
Estimation techniques for predicting population growth and fertility are discussed, and a mathematical model is developed.
Location: Princeton University Library (SPR).

57:20755 Wrigley, Neil. Unobserved heterogeneity and the analysis of longitudinal spatial choice data. European Journal of Population/Revue Europeenne de Demographie, Vol. 6, No. 4, Dec 1990. 327-58 pp. Amsterdam, Netherlands. In Eng. with sum. in Fre.
"This paper tackles the problem of handling uncontrolled heterogeneity due to unobserved influences on the decision process of spatial choice. It concentrates upon a discrete-time/'random-effects' approach to the problem of unobserved heterogeneity and documents parametric and non-parametric methods of specifying and estimating models which can cope with unobserved heterogeneity." The model is applied to data from Wales.
Correspondence: N. Wrigley, University of Wales, Department of Town Planning, P.O. Box 906, Colum Drive, Cardiff CF1 3YN, Wales. Location: Princeton University Library (SPR).

Copyright © 1991-1996, Office of Population Research, Princeton University.