Editing Demography (section)

==Methods==
[[File:1888-01-22, La Ilustración Española y Americana, El censo de población, Alcázar.jpg|thumb|Early censuses and surveys provided demographic data.]]
Demography is the [[statistical]] and mathematical study of the size, composition, and [[Population density|spatial distribution]] of human populations and how these features change over time.  Data are obtained from a census of the population and from registries: records of events like [[Birth certificate|birth]], [[death certificate|death]]s, migrations, marriages, [[divorce]]s, diseases, and [[employment]].  To do this, there needs to be an understanding of how they are calculated and the questions they answer which are included in these four concepts: [[population change]], [[standardization]] of population numbers, the demographic bookkeeping equation, and population composition.{{citation needed|date=May 2023}}

There are two types of [[data collection]]—direct and indirect—with several methods of each type.

=== Direct methods ===
Direct data comes from vital statistics registries that track all births and deaths as well as certain changes in legal status such as marriage, divorce, and migration (registration of place of residence). In developed countries with good registration systems (such as the [[United States]] and much of [[Europe]]), registry statistics are the best method for estimating the number of births and deaths.

A [[census]] is the other common direct method of collecting demographic data. A census is usually conducted by a national government and attempts to enumerate every person in a country. In contrast to vital statistics data, which are typically collected continuously and summarized on an annual basis, censuses typically occur only every 10 years or so, and thus are not usually the best source of data on births and deaths. Analyses are conducted after a census to estimate how much over or undercounting took place. These compare the [[sex ratio]]s from the census data to those estimated from natural values and mortality data.

Censuses do more than just count people. They typically collect information about families or households in addition to individual characteristics such as age, sex, marital status, literacy/education, employment status, and occupation, and geographical location. They may also collect data on migration (or place of birth or of previous residence), language, religion, nationality (or ethnicity or race), and citizenship.  In countries in which the vital registration system may be incomplete, the censuses are also used as a direct source of information about fertility and mortality; for example, the censuses of the [[People's Republic of China]] gather information on births and deaths that occurred in the 18 months immediately preceding the census.
[[File:Población Mundial.svg|thumb|Map of [[Country|countries]] by population]]
[[File:World population growth - time between each billion-person growth.svg|thumb|upright=1.6|Rate of human population growth showing projections for later this century<ref>{{cite journal | url = https://ourworldindata.org/world-population-growth#how-long-did-it-take-for-the-world-population-to-increase-by-one-billion | author1 = Max Roser | author1-link = Max Roser | author2 = Hannah Ritchie | author2-link = Hannah Ritchie | author3 = Esteban Ortiz-Ospina | author4 = Lucas Rodés-Guirao | title = How long did it take for the world population to increase by one billion? | journal = Our World in Data | year = 2013 | access-date = 25 November 2022 | archive-date = 13 October 2016 | archive-url = https://web.archive.org/web/20161013144559/https://ourworldindata.org/world-population-growth/#how-long-did-it-take-for-the-world-population-to-increase-by-one-billion | url-status = live }}</ref>]]

=== Indirect methods ===
Indirect methods of collecting data are required in countries and periods where full data are not available, such as is the case in much of the developing world, and most of [[historical demography]]. One of these techniques in contemporary demography is the sister method, where survey researchers ask women how many of their sisters have died or had children and at what age. With these surveys, researchers can then indirectly estimate birth or death rates for the entire population. Other indirect methods in contemporary demography include asking people about siblings, parents, and children. Other indirect methods are necessary in historical demography.{{citation needed|date=May 2023}}

There are a variety of demographic methods for modelling population processes. They include models of mortality (including the [[life table]], [[Gompertz curve|Gompertz models]], [[Proportional hazards models|hazards models]], [[Proportional hazards models|Cox proportional hazards models]], [[Decrement table|multiple decrement life tables]], Brass relational logits), [[fertility]] (Hermes model, [[Ansley J. Coale|Coale]]-Trussell models, [[parity progression ratios]]), marriage (Singulate Mean at Marriage, Page model), disability ([[Sullivan's method]], multistate life tables), [[population projections]] ([[Lee–Carter model|Lee-Carter model]], the [[Leslie matrix|Leslie Matrix]]), and [[population momentum]] ([[Nathan Keyfitz|Keyfitz]]).

The United Kingdom has a series of four national birth cohort studies, the first three spaced apart by 12 years: the [[National Survey of Health & Development|1946 National Survey of Health and Development]], the 1958 [[National Child Development Study]],<ref>{{cite journal |author=Power C and Elliott J |title=Cohort profile: 1958 British Cohort Study |journal=International Journal of Epidemiology |volume=35 |issue=1 |pages=34–41 |year=2006 |pmid=16155052 |doi=10.1093/ije/dyi183|doi-access=free }}</ref> the [[1970 British Cohort Study]],<ref>{{cite journal |author=Elliott J and Shepherd P|title=Cohort profile: 1970 British Birth Cohort (BCS70)|journal=International Journal of Epidemiology |volume=35 |issue=4 |pages=836–43|year=2006 |doi=10.1093/ije/dyl174 |pmid=16931528|doi-access=free }}</ref> and the [[Millennium Cohort Study]], begun much more recently in 2000. These have followed the lives of samples of people (typically beginning with around 17,000 in each study) for many years, and are still continuing. As the samples have been drawn in a nationally representative way, inferences can be drawn from these studies about the differences between four distinct generations of British people in terms of their health, education, attitudes, childbearing and employment patterns.<ref>The last three are run by the [http://www.cls.ioe.ac.uk Centre for Longitudinal Studies] {{Webarchive|url=https://web.archive.org/web/20181028075855/http://www.cls.ioe.ac.uk/ |date=28 October 2018 }}</ref>

Indirect standardization is used when a population is small enough that the number of events (births, deaths, etc.) are also small. In this case, methods must be used to produce a standardized [[mortality rate]] (SMR) or standardized incidence rate (SIR).<ref>{{cite web |url=http://www.geo.hunter.cuny.edu/~imiyares/standard.htm |title=Direct and Indirect Standardization of Mortality Rates |accessdate=26 March 2016 |url-status=dead |archiveurl=https://web.archive.org/web/20160403114527/http://www.geo.hunter.cuny.edu/~imiyares/standard.htm |archivedate=3 April 2016 }}</ref><ref>{{Cite web |url=http://www.cid.harvard.edu/cidstudents/thesis_prize/thesis_2001/appendix_a.pdf |title=examples of standardization |access-date=16 October 2022 |archive-date=7 July 2017 |archive-url=https://web.archive.org/web/20170707210711/http://www.cid.harvard.edu/cidstudents/thesis_prize/thesis_2001/appendix_a.pdf |url-status=live }}</ref>