Abstract
Following the previous recommendations and using the linear regression between the logarithms of the probabilities of dying and some palaeodemographic indicators, we propose a network of life tables specially developed for preindustrial populations, under the properties of stable populations, and discuss their advantages and drawbacks.
The first results of this work were published in 2006 and 2008 with help from Magali Belaigues-Rossard, Paul Bernier, Nadège Couvert, Carole Perraut and Arnaud Bringé (Séguy et al. 2006b, 2008). Here we present the final version of the mortality models for palaeodemographers, and a presentation currently in preparation will include variables of use to historical demographers.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
The exclusion of outlying tables is not based on a simple visual examination of the values of the studentised residuals, an inevitable source of error, but on a program that guarantees rigorous selection. Outlying observations are analysed via Residual Student with constraints fixed as a function of the value of R 2, a measure of model quality. The greater R 2, the weaker the constraints on Residual Student, so that the model quality takes account of data uncertainty, which becomes an increasingly important factor as the model is refined. The significance tests of the model’s parameters are also based on an automatic program that tests the p-value associated with each of these coefficients so that only models with a p-value below 0.05 are retained. A further constraint has been added to the model: a non-zero constant to ensure comparability between models.
- 2.
Use of logarithms is justified by the log-normal distribution of values for the juvenility index.
- 3.
Since anthropological determination of sex is still a difficult task, we present here only a sample of the modelling done on the “both sexes” sample. All the regressions are available in the additional material on the INED website.
- 4.
For women, the only possible estimate is for 5 q 5 (with R 2 = 0.71); for men, no satisfactory correlation can be found between this indicator and any of the 18 probabilities of dying (with R 2 ≥ 0.68).
- 5.
If 1 q 0 ≺ 4 q 1: the results obtained for the model parameters come very close to those found in the general model. They are only slightly better. If 1 q 0 ≺ 4 q: the model is more precise (according to the data at our disposal), but must be treated with precaution because it is based on a smaller number of tables (N = 11).
- 6.
This remark is based solely on demographically observed data, without consideration for biases that may affect the method of calculating these indicators from osteological data.
- 7.
Using an estimated probability introduces a slight bias. However, this chain estimation is restricted to the immediately preceding probability and is unlikely to distort the construction of a life table up to age 80 and over.
- 8.
It is always possible to attribute a growth rate to the youngest age classes, if it is considered that adult migration had a short-term effect on births.
References
Buchet, L., Dauphin, C., Séguy, I. (Eds.), (2006). La paléodémographie. Mémoire d’os, mémoire d’hommes (Actes des 8es journées anthropologiques de Valbonne, juin 2003). Antibes: Éditions APDCA.
Espenshade, T. J., Bouvier, L. F., & Arthur, W. B. (1982). Immigration and the stable population model. Demography, 19(1), 125–133.
Ledermann, S. (1969). Nouvelles tables-types de mortalité (coll. “Travaux et documents, Vol. 53). Paris: INED.
Poulain, M. (1981). Contribution à l’analyse spatiale d’une matrice de migrations internes. Louvain-la-Neuve: Cabay.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2013 Springer International Publishing Switzerland
About this chapter
Cite this chapter
Séguy, I., Buchet, L. (2013). Model Tables for Pre-industrial Populations. In: Handbook of Palaeodemography. INED Population Studies, vol 2. Springer, Cham. https://doi.org/10.1007/978-3-319-01553-8_8
Download citation
DOI: https://doi.org/10.1007/978-3-319-01553-8_8
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-01552-1
Online ISBN: 978-3-319-01553-8
eBook Packages: Humanities, Social Sciences and LawSocial Sciences (R0)